Google Cloud Launches BigQuery Data Agents as Virgin Media O2 Scales AI

What's new with data agents — Google Cloud Tech

Google Cloud Tech

Autonomous data agents eliminate business intelligence backlogs only when enterprise teams enforce certified data layers and automated schema context to prevent hallucinations.

AI-assisted summary of What's new with data agents — original source. Check the original for full context.

Source published:

Key ideas

  • Google Cloud reached General Availability for its Data Engineering and Data Science Agents, while reporting a 30x annual surge in data processed by Gemini in BigQuery and a 20x job increase for its BigQuery Model Context Protocol (MCP) server. (4:14)
  • The Conversational Analytics API (CA API) is now generally available, enabling organizations to embed governed conversational data agents across BigQuery, Looker, and AlloyDB into custom internal applications. (18:29)
  • Google previewed the Cloud Data Agent Kit across VS Code and terminal agents like Gemini CLI, Claude Code, and Codex, bringing native Lakehouse catalog viewers, automated pipeline generation, and model comparison directly into developer environments. (22:17)
  • Virgin Media O2 deployed its internal Gemini-powered agent platform, Parity, scaling from an initial 3-month proof of concept to spinning up certified domain agents in minutes across 750 active licenses, with plans to expand to 5,000 users. (35:23)

Detailed notes

## Background This Google Cloud Next session was published by Google Cloud Tech on YouTube on June 25, 2026. The video features speakers Ganesh Kumar Gella, Manikandan Srinivasan, and Mauro Flores exploring recent enhancements in data analytics agents, including built-in capabilities across data engineering, data science, conversational analytics, and developer tools like the Agent Development Kit and Model Context Protocol integrations.

## TL;DR The presentation details Google Cloud's vision and product roadmap for the agentic era, focusing on assistive tools like BigQuery Assistant and specialized agents for data engineering and data science. It introduces developer extensibility features, including the General Availability of the Conversational Analytics API, remote BigQuery Model Context Protocol (MCP) servers, Agent Analytics, and the Google Cloud Data Agent Kit for VS Code and terminal coding agents. Finally, Mauro Flores outlines an enterprise deployment at Virgin Media O2 (VMO2), demonstrating how their custom 'Parity' platform built on Gemini Enterprise democratized certified business analytics across hundreds of employees while cutting agent deployment time from months to minutes.

## Atmosphere & style The session carries the crisp, polished tempo of a flagship tech keynote, balancing forward-looking executive vision with pragmatic, live-terminal demonstrations. Ganesh and Mani bring structured enthusiasm and fluid product expertise, turning complex multi-step SQL optimizations and pipeline scaffolding into effortless on-stage actions. Mauro Flores injects a candid, grounded warmth to close out the talk—countering corporate PR clichés with honest reflections on organizational friction, governance discipline, and genuine end-user delight from long-tenured telecom professionals.

## Bulletin - [04:14](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=254s) Google Cloud reached General Availability for its Data Engineering and Data Science Agents, while reporting a 30x annual surge in data processed by Gemini in BigQuery and a 20x job increase for its BigQuery Model Context Protocol (MCP) server. - [18:29](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=1109s) The Conversational Analytics API (CA API) is now generally available, enabling organizations to embed governed conversational data agents across BigQuery, Looker, and AlloyDB into custom internal applications. - [22:17](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=1337s) Google previewed the Cloud Data Agent Kit across VS Code and terminal agents like Gemini CLI, Claude Code, and Codex, bringing native Lakehouse catalog viewers, automated pipeline generation, and model comparison directly into developer environments. - [35:23](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=2123s) Virgin Media O2 deployed its internal Gemini-powered agent platform, Parity, scaling from an initial 3-month proof of concept to spinning up certified domain agents in minutes across 750 active licenses, with plans to expand to 5,000 users.

## Key takeaways - [00:13](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=13s) Gartner projects that over 50% of enterprise business decisions will be automated or augmented by agentic AI by 2027. - [00:13](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=13s) Google Data Cloud structures its agentic strategy across three distinct tiers: assistive tools, specialized persona agents, and open developer protocols. - [04:14](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=254s) The Data Engineering and Data Science Agents are officially in General Availability, automating natural language pipeline generation, pipeline repair, and end-to-end model training. - [13:45](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=825s) BigQuery Assistant's integrated chat panel in BigQuery Studio automates query generation, syntax fixes, SAFE_DIVIDE wrappers, and partition pruning optimizations. - [08:04](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=484s) Conversational Analytics natively inside BigQuery and Looker enables multimodal querying, anomaly detection, and forward-looking predictive forecasting without requiring SQL knowledge. - [10:31](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=631s) Deep Data Research (Deep Dive) synthesizes structured BigQuery tables, unstructured corporate documents, and live web sources into comprehensive multi-page diagnostic reports. - [18:29](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=1109s) The Conversational Analytics API reached General Availability, allowing developers to embed governed natural language analytics into external apps and chat services. - [18:29](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=1109s) The BigQuery Agent Analytics plugin provides single-line instrumentation for agent frameworks like ADK and LangGraph to log operational telemetry and tokens directly into BigQuery. - [23:35](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=1415s) The Google Cloud Data Agent Kit unifies catalog browsing, pipeline DAG creation, and ML model comparisons directly inside VS Code and terminal CLI agents. - [31:02](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=1862s) Telecommunications provider Virgin Media O2 migrated its enterprise data footprint across 16,000 employees onto Google Cloud to establish BigQuery business layers as a single source of truth. - [35:23](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=2123s) Virgin Media O2's internal platform, 'Parity', eliminated BI reporting backlogs by enabling non-technical stakeholders to query certified domain layers via Gemini Enterprise. - [37:52](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=2272s) Mandatory operational and safety onboarding training was enforced at Virgin Media O2 as a strict prerequisite for receiving generative AI licenses. - [38:49](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=2329s) Conversational agents for telecom operations actively clarify ambiguous terms (e.g., defining what constitutes a 'call') before executing queries against production datasets. - [42:16](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=2536s) Automating YAML schema generation and embedding verified 'golden query' benchmark pairs drastically curtails agent hallucination rates across production deployments. - [44:54](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=2694s) Virgin Media O2's data agents have generated over 12,000 validated operational answers across roughly 750 active licenses, with expansion targeting 5,000 enterprise users.

## Detailed walkthrough ### [00:13](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=13s) The Dawn of the Agentic Era in Google Data Cloud

Ganesh Kumar Gella opens the session by declaring that this year has officially become "the year of agentic AI," highlighting his tremendous optimism for what lies ahead at the crossroads of data and artificial intelligence. Joining him for the presentation are Manikandan Srinivasan, Director of Product, and Mauro Flores, Assistant Vice President at VMware, who will share real-world implementation stories of using data agents within enterprise operations. Ganesh argues that the industry has fully entered an agentic era where software agents are no longer just passive tools; they are proactive, autonomous entities capable of deriving valuable business insights from massive, highly complex datasets. Backing this up with industry research, Ganesh points out:

> “Gartner is predicting that by 2027, more than 50% of business decisions are going to be augmented or automated by the agent AI.”

Ganesh emphasizes that Google is uniquely positioned to deliver on this promise because it actively innovates across every tier of generative and agentic AI, from low-level AI infrastructure all the way up to user-facing agents and applications. In the Data Cloud portfolio, agents are tailored to a wide spectrum of user personas—data analysts, data engineers, machine learning scientists, and non-technical business users.

```mermaid flowchart TD accTitle: Three Pillars of Agentic Investment in Google Data Cloud accDescr: Diagram illustrating Google Data Cloud's agentic AI tiers spanning assistive capabilities, specialized agents, and developer APIs.

subgraph S1["Assistive Capabilities"] A["BigQuery Assistant<br/>(BigQuery Studio Side Panel)"] end

subgraph S2["Persona & Task-Specific Agents"] B1["Data Engineering Agent<br/>(Pipeline Generation & Repair)"] B2["Data Science Agent<br/>(Colab Notebooks & ML Modeling)"] B3["Conversational Analytics<br/>(Looker & BigQuery)"] end

subgraph S3["Developer Tooling & Primitives"] C1["Agent Development Kit (ADK)"] C2["Model Context Protocol (MCP) APIs"] end

S1 --> S2 --> S3 ```

Google's agentic investments fall into three specific pillars: - **Assistive Capabilities:** Productive built-in tools like BigQuery Assistant integrated directly into studio environments. - **Persona- and Task-Specific Agents:** Tailored agents designed for explicit responsibilities, including the Data Engineering Agent, the Data Science Agent, and Conversational Analytics Agents. - **Developer Primitives and Open Protocols:** Foundational APIs, SDKs, and Model Context Protocol (MCP) integrations allowing developers to construct custom agents and deploy them to any endpoint or IDE.

[04:14](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=254s) BigQuery Assistant and Task-Specific Engineering & Science Agents

Focusing first on assistive features, Ganesh highlights the BigQuery Assistant, embedded in the right-hand chat panel of BigQuery Studio. The assistant empowers analysts and engineers to author SQL from scratch, troubleshoot errors, and optimize sluggish queries, enabling a workflow Ganesh refers to as "vibe coding your SQL on BigQuery Studio." The impact of these features has been dramatic: data processed by Gemini in BigQuery grew by thirty times (30x) within a single year.

Moving into dedicated persona agents, Ganesh introduces the Data Engineering Agent, targeted at eliminating the friction of data residing in disparate systems or awkward formats. Using plain natural language prompts, data engineers can draft complete, sophisticated data pipelines in minutes while remaining firmly in control to review and adjust specifications. In addition to building new flows, it analyzes and repairs broken pipelines. Ganesh follows this with the Data Science Agent, built directly into Colab notebooks in BigQuery. The agent automates data exploration, feature engineering, and end-to-end model training, letting ML scientists test accuracy metrics and validate models in minutes rather than spending weeks or months manually writing boilerplate code.

- **BigQuery Assistant:** Assists with query authoring, troubleshooting syntax/runtime errors, and identifying slow-running queries to optimize execution. - **Data Engineering Agent (General Availability):** Translates natural language requirements into robust data pipelines and debugs existing ones; officially GA. - **Data Science Agent (General Availability):** Integrated with BigQuery Colab notebooks for exploration, feature engineering, and ML model generation; officially GA.

[08:04](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=484s) Conversational Analytics Across Looker and BigQuery

Turning to the business user persona, Ganesh covers Conversational Analytics Agents, which democratize access to corporate data so non-technical decision-makers can query information and take action immediately. Looker’s conversational analytics capabilities have been in General Availability since November, featuring visualizations and interactive dashboards that allow business users to chat directly with their live dashboard tiles. Extending this functionality, Google has brought conversational analytics natively into BigQuery. Business users are no longer confined to backward-looking metrics:

> “Business users can go beyond querying the past information, and they can perform advanced analytical capabilities like forecasting, key driver analysis, or anomaly detection with this ability.”

Users can transition from asking historical questions like revenue for the prior 12 months to requesting a six-month future projection—without knowing table schemas, underlying SQL syntax, or machine learning details. The system also supports multimodal querying over images and large blocks of unstructured text, and is available in General Availability.

```mermaid flowchart LR accTitle: Conversational Analytics Publishing and Deployment Flow accDescr: Flow diagram showing how conversational agents are constructed over datasets and published to end-user surfaces.

subgraph DataSources["Data Sources"] DS1["Looker Dashboards"] DS2["BigQuery (Multimodal & Structured)"] end

subgraph AgentCore["Conversational Analytics Agent"] AG["Analysis, Forecasting & Insights"] end

subgraph EndSurfaces["Publishing Surfaces"] P1["Looker / Data Studio"] P2["Gemini Enterprise"] P3["Custom Apps via API"] end

DataSources --> AG AG --> EndSurfaces ```

[10:31](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=631s) Enterprise Publishing, Deep Data Research, and Autonomous Workflows

Once conversational agents are configured, authors need frictionless distribution channels. Organizations can publish these agents directly to Looker / Data Studio, or expose them inside Gemini Enterprise so business teams can access data agents alongside other enterprise AI agents in a single interface. Access to publishing agents into Gemini Enterprise is currently open in an allowlist preview via a QR code.

Ganesh also debuts two advanced features: 1. **Deep Data Research ("Deep Dive"):** Designed for queries that cannot be solved by structured tables alone. Deep Dive runs long-running agent investigations that combine structured BigQuery records, unstructured corporate documents, and live web information to synthesize comprehensive, narrative multi-page reports rather than shallow single-turn answers. 2. **Agentic Workflows:** Moving past pure conversational interaction, these workflows configure agents to run proactively and autonomously. They can be scheduled or wired to execute against real-time operational system events without waiting for human prompts.

Furthermore, all conversational capabilities are exposed over APIs so enterprises can embed them directly into proprietary custom user interfaces or third-party workflows.

[13:45](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=825s) Demo: Automated SQL Debugging and Slot Optimization in BigQuery Studio

The session transitions to a live demonstration inside BigQuery Studio highlighting BigQuery Assistant. The demonstrator opens an existing SQL script that fails with an error stating that `item ID` cannot be found in the referenced table. Rather than manually digging through schema definitions, the user clicks the "Fix it" button in the Gemini panel.

Gemini analyzes the query and implements multiple corrections: - Replaced the mismatched `user_id` reference with the correct `order_id` in the order items join. Noting the gravity of that misstep, the presenter remarks, “Oh, that would have been a bug.” - Fixed a mismatched `product_id` and updated references to `product_category`. - Wrapped risky division operations in `SAFE_DIVIDE` to prevent potential divide-by-zero crashes.

Running the corrected query yields successful results, but the job details show it consumed approximately 13 minutes of slot time. The demonstrator triggers Gemini’s query optimization workflow. Gemini inspects the execution plan, rewrites the date filtering conditions to enable partition pruning across the target tables, and streamlines the aggregation syntax to lower computational overhead. The demonstrator sets up a side-by-side comparison to measure the performance improvement against the baseline query.

[15:18](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=918s) Investigating Spikes in Return Rates with Gemini and Deep Dive Reports

Continuing his hands-on investigation, Ganesh Kumar Gella takes on a new task triggered by operational anomalies: a recent spike in return rates. Rather than writing queries from scratch, he asks Gemini to compose a SQL query to calculate the overall return rate across the catalog and opens the results in a new tab. Finding an overwhelming volume of raw data, he refines his prompt interactively, directing Gemini to filter the dataset to the last quarter and roll it up to the brand level.

Upon reviewing the aggregated numbers, Ganesh notices a classic edge-case distortion: several brands show a 100% return rate simply because they sold only one item and had that single item returned. To remove the noise, he instructs the model to exclude brands with 10 or fewer sales and restrict the output to the top five brands. Running the adjusted query surfaces meaningful outliers, including brands suffering from return rates as high as 41%.

```mermaid flowchart TD accTitle: Return Rate Analysis and Deep Dive Workflow accDescr: Step-by-step pipeline from natural language query generation to root cause conversational analytics. subgraph DataExploration["Iterative SQL Refinement"] Q1["Initial prompt: Return rate for everything"] --> Q2["Filter: Last quarter rolled up to brand level"] Q2 --> Q3["Sanitize: Exclude sales <= 10, show top 5 brands"] Q3 --> Outliers["Surfaced outliers (e.g., 41% return rate)"] end subgraph DeepDive["Conversational Analytics Deep Dive"] Outliers --> Prompt["Multi-factor prompt: Profitability, returns, sentiment"] Prompt --> Outline["Generated analysis objectives"] Outline --> Report["Multi-dimensional diagnostic report"] end ```

Equipped with these preliminary findings, Ganesh shifts into Conversational Analytics to perform an end-to-end root-cause analysis using the new deep dive feature. He submits a comprehensive multi-factor business prompt:

> “So I'm going to ask it which brands within our portfolio are currently at risk due to a combination of declining profitability, high return rates and negative customer sentiment. And what are the primary operational and marketing drivers behind these issues.” — Ganesh Kumar Gella

The system first provides a clear outline of its diagnostic objectives before generating a detailed diagnostic report. The resulting breakdown highlights operational and supply chain friction: - **Brand risk:** The Euro brand is singled out for abnormally high return rates. - **Supply chain bottleneck:** The Philadelphia distribution center is flagged as an operational bottleneck directly tied to elevated returns. - **Channel attribution:** The report renders an analytical breakdown correlating return volume with specific marketing acquisition traffic sources.

Ganesh summarizes the end-to-end demonstration by noting how Google Cloud's data agent capabilities streamline work “all the way from web querying to the web reporting,” before handing the stage over to Manikandan Srinivasan to cover developer-facing architectures.

---

[18:29](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=1109s) Expanding Developer Extensibility: CA API, BigQuery SDK, MCP, and Agent Analytics

Manikandan ("Mani") Srinivasan takes over to explore how developers can build and embed custom data agents. While Ganesh demonstrated first-party capabilities hosted directly inside the Google Cloud Console, enterprise customers consistently request the ability to embed those same capabilities into their proprietary web services, custom software, and internal chat applications. Addressing this requirement, Mani announces the General Availability (GA) of the Conversational Analytics API (CA API). The API incorporates enterprise governance and access control mechanisms, providing secure, governed connectivity across Google Cloud data stores—including BigQuery, Looker, and AlloyDB.

```mermaid flowchart LR accTitle: Extensibility Architecture for Custom Data Agents accDescr: Overview of CA API, BigQuery SDK, and MCP integrations connecting client applications to core data services. subgraph Clients["Custom Apps & Developer Tooling"] CA["Chat / Enterprise Apps"] SDKApp["Custom Agents (ADK / LangGraph)"] OSS["Open Source MCP Frameworks"] end subgraph MiddleTier["Integration & Agent Layer"] CA --> CAAPI["Conversational Analytics API"] SDKApp --> BQSDK["BigQuery SDK Integration"] OSS --> MCP["Managed Remote BigQuery MCP Server / MCP Toolbox"] end subgraph Storage["Google Cloud Data & Observability"] CAAPI --> Engine["BigQuery / Looker / AlloyDB"] MCP --> Engine SDKApp -.->|"1 line of code: Agent Analytics Plugin"| Telemetry["BigQuery Telemetry (Tokens, Latency, Logs)"] end ```

Mani outlines several major GA milestones expanding developer access across custom codebases and open protocols: - **BigQuery SDK Integration:** Enables developers authoring agents via the Google SDK framework to directly leverage built-in BigQuery toolsets. - **BigQuery Model Context Protocol (MCP) Server:** A fully managed remote MCP server providing complete BigQuery access without requiring developers to write brittle custom glue code or API connectors. - **MCP Toolbox:** Designed for engineering teams relying on open-source agentic frameworks who prefer managing their own local or self-hosted MCP tool environments. - **Rapid Adoption Metrics:** BigQuery MCP adoption has seen explosive growth over the prior six months, processing millions of jobs and achieving a 20x year-over-year increase in job volume.

Turning to day-two operations, Mani highlights real-time agent observability. When running agents in production, teams must monitor operational health, token usage patterns, latency profiles, and qualitative end-user sentiment. To solve this, Mani introduces the General Availability of the BigQuery Agent Analytics plugin. With a single line of instrumentation code, developers using frameworks like the Agent Development Kit (ADK) or LangGraph can log all interaction telemetry, execution metrics, and performance signals directly into BigQuery tables. As an architectural tip, Mani suggests bootstrapping Conversational Analytics directly on top of those agent analytics tables, allowing engineers to audit agent runtime health and debug production behavior using plain natural language.

---

[22:17](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=1337s) Google Cloud Data Agent Kit and VS Code Integration

Mani transitions to address developer ergonomics across modern engineering teams. Data practitioners, analytics engineers, and data scientists operate across fragmented environments—some working inside graphical IDEs like VS Code, while others favor terminal-based coding agents such as Claude Code, Gemini CLI, or Codex. Historically, integrating enterprise data tools required developers to hand-craft bespoke skills, custom connectors, and local MCP configurations for each individual environment.

To unify and automate these workflows, Mani recaps the preview launch of the Google Cloud Data Agent Kit, delivered across two core formats: 1. **A VS Code Extension:** Available directly via the Visual Studio Code Marketplace for interactive development within the IDE. 2. **Terminal Agent Plugins:** Native plugins purpose-built for Gemini CLI, Claude Code, and Codex.

Emphasizing Google's open approach, Mani points out that the underlying skills and Model Context Protocol tools are open source, allowing engineering teams to audit, modify, and extend them to suit custom organizational environments.

> “And we are very confident that this is going to rapidly accelerate developer velocity.” — Manikandan Srinivasan

---

[23:35](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=1415s) Demo: Onboarding, Data Exploration, and Automated Pipeline Generation

Mani walks through a live demonstration of the Data Agent Kit in VS Code, starting with its onboarding flow. Upon installing the extension, users navigate a guided configuration pane where they can inspect and connect their choice of managed or local MCP servers. Once the protocol bridge is configured, developers select their target datasets and access a suite of predefined, open-source skills that can be reviewed and edited directly within the editor.

The extension functions as a centralized control plane for data exploration, packaging deep ecosystem visibility into dedicated native viewers: - **Catalog Viewer:** Unifies access across Lakehouse storage, BigQuery tables, AlloyDB, and Cloud Spanner, complete with fast search across schemas. - **Storage & Job Inspect:** Direct views into Google Cloud Storage (GCS) objects, historical and active BigQuery job executions, and saved SQL queries. - **Spark & Pipeline Viewer:** Gives data engineers direct insight into operational data pipelines and Apache Spark executions straight from the IDE panel.

Mani then demonstrates the natural language data chat integrated into the extension, powered directly by Conversational Analytics. Instead of writing boilerplate queries, developers can query datasets using ordinary phrasing—such as asking for Q4 revenue figures or exploring table profiles—and receive tabular results alongside auto-generated visual charts without leaving their editing environment.

To illustrate more sophisticated workflows, Mani initiates an automated fraud detection use case. Rather than requiring manual scripting, the Data Agent Kit takes a plain prompt, analyzes the available dataset metadata, and formulates an end-to-end execution plan using its configured skills. As code streams dynamically into the editor window, the agent cleanses the dataset, executes feature engineering, and synthesizes the entire operational data engineering pipeline automatically. Developers simply provide high-level intent, and the agent takes care of the implementation details.

[26:48](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=1608s) Streamlining Pipeline Creation, Testing, and Model Comparison

The live demonstration wraps up by detailing how the Agent Development Kit simplifies end-to-end data engineering and data science workflows directly inside the developer's environment. The tool does not simply output pipeline code; it renders the complete Directed Acyclic Graph (DAG) illustrating exactly how each pipeline component is assembled. Developers can execute and test their pipelines directly from the single interface without switching across multiple disparate dashboards or tools.

Once the pipeline execution finishes and verifies successfully, developers can jump immediately into machine learning workflows. Users can issue a single prompt instructing the agent to build or compare multiple models. The Agent Development Kit draws upon built-in data science skills to generate all the necessary scaffolding code and returns comparative evaluation metrics directly inside the IDE. Ganesh Kumar Gella summarizes the broader impact before handing over the stage: “So to summarize, we've actually vastly accelerated the developer productivity using data Agent Kit.”

```mermaid flowchart LR accTitle: Agent Kit Pipeline and ML Workflow accDescr: The workflow from prompting the Agent Kit to pipeline execution and comparative model evaluation within a single IDE interface.

subgraph Pipeline["Data Engineering"] A["Single User Prompt"] --> B["Agent Kit / Data Science Skills"] B --> C["DAG Generation & Scaffolding"] C --> D["Inline Pipeline Execution & Test"] end

subgraph ML["Model Evaluation"] D --> E["Model Creation & Comparison Prompt"] E --> F["Inline Model Metrics in IDE"] end ```

- Developers can view pipeline DAG assembly and run validation tests directly inside the unified interface without opening outside tools. - Data science skills embedded within the Agent Development Kit automate scaffolding generation for rapid model creation. - Multi-model comparisons and metric evaluations are rendered natively within the IDE from a single conversational prompt.

---

[28:10](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=1690s) Introducing Mauro Flores and Virgin Media O2's Scale

Ganesh invites Mauro Flores to the stage to present a real-world enterprise adoption story. Acknowledging his team in the audience and joking about the pressure of speaking last, Flores stresses that his presentation will avoid corporate public relations spin in favor of an honest breakdown of real implementations, tangible successes, and difficult challenges. Originally from Mexico and based in the UK for many years, Flores brings over 25 years of technology experience, spending the last three years in telecommunications—an industry historically dominated by customer churn concerns. Alongside his tech career, Flores has practiced martial arts for over three decades, though he emphasizes that his identity is rooted first and foremost in being a husband and father.

Flores explains the scale of Virgin Media O2 (VMO2), the UK's second-largest telecommunications and media provider, which was created by combining mobile and broadband networks. Handling roughly 45 million connections across broadband, mobile, and television services, VMO2 covers approximately 99% of the UK population on mobile and is driving an aggressive fiber-first deployment targeting 80% coverage by 2028 alongside net-zero sustainability goals. Flores notes that while telco used to focus almost exclusively on physical infrastructure—laying cables and mounting antennas—modern customer expectations require software, data, and agile digital technology to run the business.

- Mauro Flores leads data initiatives at Virgin Media O2 (VMO2), leveraging 25+ years in technology and 30+ years in martial arts. - VMO2 manages roughly 45 million connections spanning mobile, broadband, and television across the UK. - The company provides 99% mobile network coverage across the UK population and is targeting 80% fiber-first footprint completion by 2028 alongside net-zero targets.

---

[31:02](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=1862s) Migrating to Google Cloud and Establishing BigQuery Business Layers

Historically, telecom companies operated as passive utilities: a customer moved into a home, plugged in their router or mobile device, and paid their monthly bill. Flores explains that transforming this massive, rigid operational machine into an agile, customer-centric organization required a modern data strategy. Over the preceding two and a half years, VMO2 embarked on a company-wide migration to Google Cloud Platform (GCP) across 16,000 employees, consolidating systems from distinct historical entities onto BigQuery.

Flores underscores that cutting-edge AI tools are completely ineffective if fed bad data: “Again rubbish in, rubbish out.” To prevent this, VMO2 established BigQuery as their absolute single source of truth, structuring it into curated business layers tied to specific domains. Despite this clean data foundation, traditional BI and self-service dashboards created a significant bottleneck. Central teams faced never-ending sprint backlogs while building dashboards and visualization assets. Moreover, asset discovery became an uphill battle, limiting how quickly business users could turn data into decisions that improved customer retention and revenue.

- VMO2 migrated disparate legacy systems across a 16,000-employee enterprise onto a centralized Google Cloud Platform environment. - BigQuery was architected as the single source of truth, organized into certified business layers aligned with discrete operational domains. - Traditional dashboard self-service created massive backlogs, impeding rapid data democratization and discoverability.

---

[33:39](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=2019s) Conversational Analytics to Empower Curiosity Across 800 Data Professionals

Seeking to eliminate self-service bottlenecks, VMO2 partnered early with Google to develop conversational analytics capabilities. Across the wider organization, approximately 800 people work in data-related roles. Rather than spending their days answering repetitive metric requests, these specialists needed to be freed up for higher-value initiatives. The primary objective was providing non-technical stakeholders direct, natural-language access to certified metrics without intermediary support tickets.

Flores emphasizes that conversational analytics must do more than answer basic transactional questions like daily sales totals; it must stimulate deeper user curiosity. Technology that looks impressive during demos is useless if it fails to yield actionable business outcomes. Flores highlights this distinction sharply: “Don't you get normally people we're talking about great tech, great tech without deriving value means nothing.” By enabling stakeholders to ask spontaneous follow-up questions and explore data organically, the organization aims to transform how business teams analyze problems and discover growth opportunities.

- VMO2's extended data community encompasses roughly 800 specialists whose time was historically consumed by routine data requests. - Conversational analytics was designed to enable direct, unassisted stakeholder queries against curated business data. - The business objective centers on promoting iterative data exploration and genuine commercial value rather than deploying technology for its own sake.

---

[35:23](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=2123s) Building "Parity": Enterprise Agent Architecture on Gemini

To implement conversational analytics safely, Flores's engineering team built an internal platform named "Parity" on top of Google's conversational frameworks and powered by Gemini Enterprise. The biggest challenge with data agents in high-stakes enterprise settings is maintaining strict factual consistency. If a business leader queries critical key performance indicators—such as top-line revenue—and receives conflicting numbers across different queries, trust in the system collapses. Parity solves this by routing all conversational queries exclusively through certified BigQuery business layers via an analytics API, complete with features to store recurring queries as favorites.

```mermaid flowchart TD accTitle: Parity Agent Architecture accDescr: Architecture of the Parity platform showing certified BigQuery layers routed through an Analytics API to Gemini Enterprise, with governance controls and feedback loops.

subgraph DataFoundation["Certified Foundation Layer"] A["BigQuery Single Source of Truth"] --> B["Certified Business Domain Layers"] end

subgraph CoreEngine["Agent & API Orchestration"] B --> C["VMO2 Analytics API"] C --> D["Parity Agent Platform (Gemini Enterprise)"] D --> E["User Interface & Saved Favorites"] end

subgraph GovernanceFeedback["Control & Improvement"] D --> F["Governance, Privacy & Legal Gates"] D --> G["Continuous Feedback & Evaluation Loops"] G -.->|"Ongoing Model & Prompt Tuning"| D end ```

Flores highlights the dramatic acceleration of their development lifecycle: while their initial proof-of-concept took three months of exploratory effort to get production-ready, the mature framework now allows the team to spin up new, fully functional domain agents in a matter of minutes. Flores jokes about the reaction from internal colleagues: “Now if someone from VMO2 listening to this in days. Yeah, but minutes. I'm not. I'm not the PR guy. It's minutes.” However, he cautions that rapid technical deployment is only possible when foundational data quality, privacy controls, legal compliance, and governance frameworks are established in advance. The architecture is rounded out with rigorous continuous-feedback loops that capture interactions and maintain answer consistency over time, forming the basis for rapid prototyping and enterprise empowerment.

- The Parity agent platform is built on Gemini Enterprise, pulling from certified BigQuery business layers via an internal Analytics API to ensure consistent revenue metrics. - Agent creation time dropped from a three-month initial proof-of-concept phase to deployment within mere minutes once standardized foundational layers were in place. - Robust data governance, privacy compliance, and systematic feedback loops prevent hallucinations and regulatory roadblocks across production deployments.

[37:52](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=2272s) Securing Buy-In, Procurement, and Mandatory User Training

In technical and engineering environments, adopting generative AI tools often triggers hesitation and skepticism around whether the systems will actually perform reliably in production. Securing early business alignment and rapid procurement approvals was essential to overcoming these barriers. The engineering leadership adopted an assertive posture with procurement partners to accelerate the onboarding process while guaranteeing that governance and data safety standards remained uncompromised. To build credibility and satisfy internal compliance demands, the team introduced a non-negotiable policy: access to conversational analytics tools is conditioned upon formal safety and operational training.

Every employee assigned a license must successfully complete onboarding and training protocols before accessing the platform; failure to comply results in immediate revocation of their access credentials. This strict governance model laid the operational foundation for scaling the tool across hundreds of employees without exposing sensitive corporate data to unvetted usage patterns. As Mauro Flores put it: “So everyone that goes into one of the licenses that we have, they get onboarded and they have to get trained.”

- **Stakeholder Strategy:** Aggressive alignment with procurement teams to bypass approval bottlenecks while ensuring strict enterprise safety controls. - **Access Governance:** Completion of conversational analytics training is strictly mandatory; uncompleted training leads to immediate tool license forfeiture.

---

[38:49](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=2329s) Live Demo: Conversational Analytics on Customer Interaction Data

Demonstrating the practical value of these agents, Mauro Flores walked through a customer contact center use case common across telecommunications providers, where subscribers frequently reach out regarding mobile or broadband service issues. Interacting with customer contact teams constitutes the highest-frequency customer touchpoint for the telco, making customer contact operations a prime target for AI-driven analytics. Accessing the customer interaction agent through the enterprise portal, an analyst can initiate inquiry using everyday business language, such as asking for total call volumes over the past few months. While seemingly straightforward, the concept of a "call" carries multiple conflicting definitions across business units. Rather than hallucinating or producing an arbitrary calculation, the agent queries the underlying BigQuery business semantic layer and actively asks the user for clarification regarding the specific definition of "call" they intend to measure.

```mermaid flowchart TD accTitle: Conversational Analytics Agent Interaction Flow accDescr: Step-by-step interaction between an enterprise user, the conversational agent, and BigQuery.

subgraph UserInteraction ["User Interface"] A["User submits prompt: 'How many calls in last few months?'"] C["User specifies context & requests channel breakdown"] E["User switches view to line graph & saves to Favorites"] end

subgraph AgentLogic ["Conversational Agent"] B["Agent seeks disambiguation: 'What do you mean by call?'"] D["Agent validates assumptions, calculates results & explains methodology"] end

subgraph DataLayer ["Enterprise Data"] F[("BigQuery Business Layer")] end

A --> B B --> C C --> D D <--> F D --> E ```

Once the user provides disambiguating context, the agent retrieves the verified figures and supports intuitive follow-up queries, such as breaking down the volume across communication channels like voice calls and Messenger. Instead of forcing users to sift through complex dashboards or build custom drill-downs, the agent prompts the user to verify channel criteria, validates the query logic, and generates transparent explanations showing exactly how the numbers were calculated. Users can dynamically switch visualizations—such as rendering trends as a line graph—and save frequently executed prompts into a personal "Favorites" list that supports on-demand data refreshes. Reflecting on the transparency challenges inherent in conventional reporting, Mauro Flores observed: “Big part of the advantage of these agents as well is we are able to go back and explain it.”

- **Ambiguity Resolution:** The agent actively clarifies business terminology with the user before compiling queries against BigQuery. - **Multichannel Drill-Down:** Dynamic segmentation across voice calling and messaging platforms to assist resource deployment and operational scheduling. - **Explainability & Customization:** Provides step-by-step logic detailing how calculations were performed, along with flexible chart re-rendering (e.g., bar to line charts). - **Workflow Persistence:** Built-in favorites functionality allows analysts to store, rerun, and refresh recurring operational queries.

---

[42:16](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=2536s) Building Agents with Automated Context Generation and Golden Queries

To democratize agent creation beyond specialized engineering teams, the organization developed an internal self-service portal directly on top of Google Cloud's data agent APIs. When a business analyst requests a specialized agent, the creator provides a recognizable agent name and selects the specific data tables authorized for exposure. In the background, a dedicated secondary agent inspects the selected schemas, extracts metadata, and automatically constructs the detailed YAML configuration files required by the platform APIs. This backend automation enriches field-level context and property descriptions across the schema, drastically reducing the hallucination rates that typically emerge when schema descriptions are incomplete or manually drafted.

```mermaid flowchart LR accTitle: Automated Agent Generation Pipeline accDescr: Workflow showing automated schema profiling, YAML compilation, and Golden Query tuning.

subgraph ConfigStage ["Configuration & Enrichment"] T["Select Tables"] --> BG["Background Agent"] BG -->|"Auto-generates"| Y["Detailed YAML Schema & Field Context"] end

subgraph TuningStage ["Evaluation & Tuning"] Y --> GQ["Inject Golden Queries (Expected Q&A pairs)"] GQ --> TST["In-App Interactive Testing"] end

subgraph DeployStage ["Deployment"] TST --> DEP["Deploy to Gemini Enterprise with Access Controls"] end ```

To guarantee deterministic, production-grade output, the workflow incorporates "golden queries"—predefined pairs of standard natural language questions and their verified, expected SQL queries. By baking golden queries into the agent's property files, the engineering team effectively tunes the model to execute vetted queries when corresponding business questions are posed. Creators can interactively run test prompts directly within the management console to confirm precision before deployment. Once verified, the agent is published to Gemini Enterprise, where granular identity and access management controls govern departmental access. Emphasizing the criticality of automated field-level enrichment, Mauro Flores stressed: “This is where if you have the right context, then this works well. Where you don't have the right context then this completely fails.”

- **Schema Profiling Agent:** A background agent automatically generates YAML configuration files and populates detailed semantic descriptions field by field. - **Golden Query Calibration:** Incorporates verified question-and-query benchmarks to ensure accurate, repeatable SQL execution for mission-critical inquiries. - **Testing & Deployment:** Live sandboxed testing environment enables verification prior to publishing the agent directly into Gemini Enterprise.

---

[44:54](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=2694s) Enterprise Impact, Scaling Milestones, and Closing Remarks

The real-world business reception to these conversational analytics agents has surpassed expectations across long-standing operational teams. Users with decades of experience in the telecommunications industry have reached out directly to express their enthusiasm, praising the agent as an unprecedented breakthrough for internal tooling. One senior employee with over twenty years of service at Virgin Media O2 (VMO2) emailed leadership directly to state that the conversational agent represented the single best internal platform ever built at the company.

The program is executing a rapid enterprise expansion roadmap. While the deployment is still in its initial rollout phase, it has already scaled to approximately 700 to 750 active licenses, which have generated over 12,000 verified operational answers. The leadership team has committed to expanding the platform to 5,000 licensed users across the enterprise by the end of the year. Concluding the session, Ganesh Kumar Gella thanked the audience, reiterating the transformative potential demonstrated by autonomous data analytics agents and encouraging attendees to participate in upcoming deep-dive technical sessions across Google Cloud Next. Recalling user feedback from long-tenured personnel, Mauro Flores shared: “I got an email from one of the users that said, people that have been in telco for 20 plus years as well, right: ‘This is the best tool I've ever seen built in VMO2.’”

- **Current Adoption:** Between 700 and 750 enterprise licenses currently active, delivering more than 12,000 automated answers. - **Scale Target:** On track to reach 5,000 active licensed enterprise users by the end of the calendar year. - **User Validation:** Unsolicited feedback from 20-year telecommunications veterans citing the platform as the benchmark for internal tooling.

## Notable quotes > [00:13](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=13s) Gartner is predicting that by 2027, more than 50% of business decisions are going to be augmented or automated by the agent AI. — Ganesh Kumar Gella

> [08:04](https://www.youtube.com/watch?v=Z-AfOcWO_kk&t=484s) Business users can go beyond querying the past information, and they can perform advanced analytical capabilities like forecasting, key driver analysis, or anomaly detection with this ability. — Ganesh Kumar Gella

## Mentioned AlloyDB, Apache Spark, BigQuery, BigQuery Assistant, BigQuery Studio, Claude Code, Cloud Spanner, Codex, Colab, Conversational Analytics API, Ganesh Kumar Gella, Gartner, Gemini CLI, Gemini Enterprise, Google Cloud Data Agent Kit, Google Cloud Platform, Google Cloud Storage, LangGraph, Looker, Manikandan Srinivasan, Mauro Flores, Model Context Protocol (MCP), Parity, Virgin Media O2 (VMO2), VMware, VS Code

## Video description Discover the latest in data analytics agents. In this session, we’ll explore enhanced built-in capabilities designed to streamline workflows across data engineering, data science, and conversational analytics. Learn to build custom agents grounded in your enterprise data using new developer tools, including the Agent Development Kit (ADK) and Model Context Protocol (MCP) integrations. Find out how to access these powerful features seamlessly via the Gemini CLI, Claude Code, Codex, and IDEs like VSCode to supercharge your productivity.

Watch more Google Cloud Next 2026 Sessions → https://goo.gle/gcloud-next-2026 Subscribe to Google Cloud Tech → https://goo.gle/GoogleCloudTech

Speakers: Manikandan Srinivasan, Ganesh Kumar Gella, Mauro Flores

BRK2-037 #GoogleCloudNext

[Source](https://www.youtube.com/watch?v=Z-AfOcWO_kk)

Reading summary

Google Cloud is expanding its Data Cloud portfolio with specialized, autonomous data agents alongside open developer primitives, while enterprise adopters report rapid deployment cycles when agents are tethered to curated semantic layers. Persona-Specific Agents and In-Studio Assistance (4:14) Google Cloud organizes its agentic investments across assistive tools, persona-specific agents, and developer primitives. Ganesh Kumar Gella highlights that Gemini-processed data in BigQuery grew 30x in one year, supported by BigQuery Assistant's SQL generation, error debugging, and execution plan optimization like partition pruning. For data teams, Google introduced generally available Data Engineering and Data Science Agents to generate data pipelines, repair broken flows, and automate model training within BigQuery Colab notebooks. Conversational Analytics and Multi-Source Investigations (8:04) Conversational Analytics agents have expanded natively into BigQuery and Looker, enabling non-technical users to run forward-looking projections, anomaly detection, and multimodal queries over unstructured data. For broader strategic investigations, Google introduced Deep Data Research ('Deep Dive'), which runs multi-turn agent workflows combining structured BigQuery tables, unstructured internal documents, and live web research into synthesized narrative reports. Agents can also run autonomously against scheduled operational triggers without waiting for manual human prompts. Extensibility, Open Protocols, and Developer Tooling (18:29) To support custom enterprise integration, Google released the Conversational Analytics API into General Availability alongside built-in BigQuery SDK tools and a managed Model Context Protocol (MCP) server. Manikandan Srinivasan noted that BigQuery MCP adoption increased job volume 20x year-over-year. Developers can also instrument real-time observability using the BigQuery Agent Analytics plugin with one line of code in frameworks like ADK or LangGraph. In addition, the preview Google Cloud Data Agent Kit delivers open-source skills and native viewers across VS Code and terminal coding agents. Enterprise Implementation and Governance at Virgin Media O2 (35:23) Virgin Media O2 built an internal conversational agent platform named 'Parity' on Gemini Enterprise, connecting directly to certified BigQuery semantic layers via an analytics API. Assistant VP Mauro Flores explained that after establishing standardized business data layers and automated schema profiling, new domain agent deployment time dropped from three months for the initial proof-of-concept to mere minutes. To protect data accuracy and maintain user trust, the platform resolves metric ambiguities through conversational clarification, incorporates predefined 'golden queries' to anchor SQL generation, and mandates formal compliance training before granting user licenses. Virgin Media O2 has scaled its conversational data platform to over 700 active licenses and 12,000 verified answers, targeting 5,000 users by year's end. However, enterprise leaders emphasize that successful autonomous data agents depend entirely on prerequisite investments in clean data architecture, strict semantic governance, and continuous human feedback loops.

Source quotes

[0:13] Ganesh Kumar Gella: Gartner is predicting that by 2027, more than 50% of business decisions are going to be augmented or automated by the agent AI. [8:04] Ganesh Kumar Gella: Business users can go beyond querying the past information, and they can perform advanced analytical capabilities like forecasting, key driver analysis, or anomaly detection with this ability.

Original transcript

[0:13] Hello, everyone. [0:16] Today, I'm super excited to talk about what's [0:20] new with the data agents. [0:23] As we all have seen, this year has been the year of agentic AI, [0:30] and I am super optimistic about what's ahead of us [0:34] at the intersection of data and AI. [0:38] I am Ganesh Geetha, senior director [0:40] of engineering for BigQuery AI. [0:43] Today I am accompanied by my colleague Manish Srinivasan who [0:49] is director of product and we are also [0:52] happy to have Mauro Flores, who is AVP from VMware two, [0:57] and he's going to talk to us about how they're using our data [1:01] agents in their organization. [1:04] With that, let's dive in. [1:08] As you see here, we are truly in the agentic era today. [1:14] In this agentic era, these agents [1:16] are capable of providing valuable insights out [1:21] of massive scales of data and the data that [1:24] is super complex to comprehend. [1:28] These agents are not only capable of providing insights, [1:32] but they are also proactive and autonomous. [1:38] This is not just me saying as you see this quote here. [1:42] Gartner is predicting that by 2027, more than 50% [1:48] of business decisions are going to be augmented or automated [1:54] by the agent AI. [1:56] It's only 2027, not too far away and only next year. [2:03] And here at Google, we are strategically positioned [2:08] to deliver this full promise to our customers [2:11] because in Google, we are innovating [2:15] at all the layers of generative AI and agentic AI, [2:20] starting from the bottommost layer of AI infrastructure [2:25] to all the way to the top layer of agents and applications. [2:33] And in Google Data Cloud, we have [2:37] agents and agentic capabilities for every data user. [2:44] The data users of data analytics are [2:47] broad and come with varied experiences [2:52] like data analysts, data engineers, machine learning [2:57] scientists, ML engineers or business users, [3:02] and we have developed and designed our agents [3:06] to meet the needs of this broad spectrum of persona. [3:12] At a high level, we categorize our agentic investments [3:17] into these three areas. [3:19] We started with assistive capabilities, [3:23] where you find these agents and agents and applications [3:27] available on our Studio and helps you to become [3:31] more and more productive. [3:33] And the second area is our data agents such as data engineering [3:38] agent, data science agent, or conversational analytics agents. [3:43] These agents are persona specific and task specific. [3:49] And the third area of investment is [3:53] on providing primitive APIs and tools such as SDK tools or MCP [3:58] APIs, so that savvy developers can build their own agents [4:04] and deliver them on a surface of their choice. [4:09] So let's start diving into these. [4:12] Category by category. [4:14] So in the assistive area I would like to introduce our BigQuery [4:19] assistant. [4:21] Our BigQuery assistant is available on BigQuery Studio. [4:25] You find it on the right hand side chat panel [4:29] and helps data engineers and data analysts [4:33] with authoring SQL. [4:35] A brand new SQL or even a troubleshoot an existing SQL. [4:40] With BigQuery assistant by your side, [4:43] you can actually start vibe coding your SQL on BigQuery [4:46] Studio, and our BigQuery assistant not only [4:51] helps with authoring the SQL or are troubleshooting the SQL, [4:54] but it also helps you to optimize the existing equals, [4:58] or even find out which of your queries [5:00] are running late and optimizing them as well. [5:04] And we are adding more and more capabilities [5:06] to the BigQuery assistant. [5:09] And because of our breadth and depth [5:12] of capabilities that we are adding, [5:15] the data processed by Gemini in BigQuery [5:18] grew up by 30 times in a single year. [5:23] Now moving on to the persona and task specific agents, [5:29] let me introduce our data engineering agent. [5:34] All of us want to derive business insights with our data. [5:40] But many a times this data is not [5:43] present in a single location, even if it is [5:46] present in a single location. [5:49] The format and the shape of the data [5:52] is not the way we wanted it to be. [5:54] And this is where data engineers build data pipelines. [5:59] And with our data engineering agent, [6:02] it helps data engineers to build sophisticated pipelines [6:06] in a matter of minutes, starting with simple natural language [6:10] and in all the way through throughout the journey. [6:13] Data engineers are in their control. [6:16] They can review the specifications [6:18] provided by the agent and modify the plans [6:21] so that they can arrive at data pipelines that [6:25] meet the business needs. [6:29] Our data engineering agent also helps [6:32] with troubleshooting the existing pipelines [6:34] and fixes them as well. [6:37] And today, I'm pleased to announce [6:40] that our data engineering agent is available in GA. [6:47] Thank you. [6:49] And moving on to another task specific agent [6:59] is our data science agent. [7:02] Our data science agent comes up on Colab notebooks of BigQuery [7:07] and helps data scientists to build sophisticated machine [7:11] learning models, starting with simple natural language. [7:14] Our data science agent knows how to explore the data [7:19] and featurize the data, and build machine learning [7:22] models that are needed for the business outcome. [7:26] Using the data science agent, machine learning scientists [7:30] now can validate the models, test them [7:33] and identify the accuracy scores for the models that [7:36] are generated by this data science agent. [7:40] Tasks that used to take weeks and months [7:44] now can be accomplished with a matter of minutes. [7:47] Any of you here data scientists and have used our data science [7:52] agent. [7:54] It's so breeze and the tasks can be completed [7:57] in a matter of minutes. [7:58] Again, this data science agent is also available in GA. [8:04] Now moving on to the third category of the agent [8:09] is our conversational analytics agents. [8:13] This is catered towards business user persona [8:17] and aims to democratize the access [8:21] to the data with conversational analytics agent. [8:25] Business users can obtain the answers to their questions [8:30] so quickly, so that they can actually [8:33] use that time savings to actually act on the information [8:36] that they are getting from conversational analytics agent. [8:40] Anyone here used conversational analytics agent of BigQuery. [8:45] OK, some of you have used them. [8:47] That's good to know. [8:51] And our conversational analytics agent [8:54] has been available on Looker and been in GA since November. [9:00] Conversational analytics agent in Looker [9:02] comes with rich visualizations, and most recently, we also [9:09] added this ability on dashboards so [9:12] that when you are on Looker dashboards, [9:14] you can start chatting with the dashboards [9:17] and start getting the rapid answers. [9:22] And now we also have conversational analytics [9:25] directly working on BigQuery with the conversational [9:30] analytics in BigQuery. [9:32] Business users can go beyond querying the past information, [9:38] and they can perform advanced analytical capabilities [9:41] like forecasting, key driver analysis, or anomaly detection [9:47] with this ability. [9:49] Business users not only can ask questions like, hey, [9:53] what was my revenue in the last 12 months. [9:55] But they can also start asking questions like, hey, [9:57] can you predict. [9:58] Can you project my revenue for the next six months, [10:01] all without having to know anything [10:04] about the tables or the schema or the models behind it. [10:09] Conversational analytics in BigQuery [10:12] also enables multi-modal data processing [10:16] so that complex large text or images, all of those [10:20] now can be queried with a simple natural language. [10:24] And our conversational analytics in BigQuery [10:27] is also available in GA. [10:31] So some of you said you have used conversational analytics [10:35] agents. [10:36] So once agent owners build these agents, [10:40] you want a surface where you can distribute them [10:43] to your business persona. [10:44] And that is where we have enabled publishing [10:47] of these agents to Data Studio. [10:50] And here you can now give access to your broader set of business [10:54] users on the similar lines. [10:58] We now added the capability to publish these data [11:03] conversational agents onto Gemini enterprise as well. [11:07] So with that business, users have one stop place [11:10] to exercise their data agents alongside other agents. [11:15] This feature is brand new and available on allowlist preview. [11:21] You can scan the QR code if you would like [11:23] to get access to this feature. [11:29] And now let me introduce another brand new feature [11:35] that we have recently added, which is our deep data research [11:39] or called deep dive. [11:42] This feature is available on Gemini enterprise, [11:46] and where it helps is a lot of times our answers that we need [11:52] is not only present in the structured data, [11:56] we need information synthesized from unstructured and even [12:01] from real time web information and from our structured sources [12:05] like BigQuery. [12:06] This is where deep data research combines these multiple sources [12:12] and provides comprehensive reports. [12:15] What deep data research does is it goes beyond pointed questions [12:19] and answers, but comes up with a long and comprehensive report. [12:23] And the agent also is capable of running [12:26] for longer periods of time. [12:29] And one more brand new feature that [12:32] is in preview that I would like to introduce [12:34] is our agentic workflows. [12:37] As we have spoken, our agents in the agentic era [12:40] are proactive and autonomous. [12:43] So with agentic workflows, now you can configure the workflow [12:47] and have it triggered on an event or on a scheduled basis. [12:55] As we have seen, conversational analytics [12:59] works across products like Looker BigQuery databases [13:05] and these agents can be published to various surfaces [13:09] like Data Studio or Gemini enterprise. [13:14] And this agent is also available via API [13:17] so that agent owners can take them and deploy them [13:21] on a surface of your choice. [13:25] So with that, now let's take a look at these agent capabilities [13:30] using this demo. [13:32] And in this demo, we are going to show our BigQuery assistant [13:35] on BigQuery Studio. [13:37] And then we are going to show the deep dive aspect [13:40] of conversational analytics. [13:45] So I have a couple of tasks that I'm going to do today. [13:48] The first one is that I have this query with an error. [13:51] We see that item ID is not found inside this table. [13:55] So I'm just going to ask Gemini app to fix this [13:58] by clicking fix it. [14:03] We sped up the video and it found a couple of fixes. [14:07] So the first thing is it fixed my order items order [14:11] ID which was user ID. [14:14] Oh, that would have been a bug. [14:17] Let's see. [14:17] It also updated the original product ID 1. [14:24] It fixed product category. [14:28] It adds safe divide. [14:29] All right. [14:32] Let's go ahead and run that. [14:39] So that looks good. [14:42] I look at my details. [14:45] Yeah took about 13 minutes of slot time. [14:49] I wonder if Gemini can optimize it for me. [14:52] So I'm going to ask Gemini to optimize. [14:59] So the first thing it did was it updated my date. [15:04] This will allow for partition pruning. [15:07] It also added a cleaner aggregation. [15:11] I want to run these side by side just to see. [15:18] All right. [15:19] Definitely an improvement. [15:23] So I have another task. [15:30] We recently saw a spike in return rates [15:33] and I need to investigate. [15:35] So I'm going to ask Gemini to write a new SQL query to find [15:39] the return rate for everything. [15:45] All right, so put together a query for us. [15:49] And let's go ahead and run it. [15:52] So I'll open this in a new tab. [16:01] And it looks like there's a lot of results here. [16:05] So let me go ahead and filter this to just the last quarter [16:10] and roll it up to the brand level. [16:16] All right. [16:23] So looking at this, I see a lot of these had one item [16:29] and returned one item that gives them a 100% return rate. [16:33] It's not super useful. [16:35] So let's go ahead and filter out any brands [16:38] that sold 10 or fewer items and just show me [16:41] the top five brands. [16:45] All right. [16:46] Let's go ahead and run that. [16:50] Great I can see a couple of these [16:53] have had really high return rates. [16:55] 41% in this case. [16:58] So I think I'm ready for my report. [17:03] All right. [17:04] And now I can do my report in conversation analytics using [17:08] the new deep dive feature. [17:10] So I'm going to ask it which brands within our portfolio [17:13] are currently at risk due to a combination of declining [17:16] profitability, high return rates and negative customer sentiment. [17:20] And what are the primary operational and marketing [17:22] drivers behind these issues. [17:27] The first thing it's going to do is give me [17:29] an outline of its objectives. [17:32] So go ahead and start the analysis. [17:42] Let's look at this report now. [17:45] So it gives me a report on my high return rates [17:48] and my operational bottlenecks. [17:51] If I skim through here, I can see it gave me a chart with. [17:56] It shows me that my Euro brand has a really high return rate. [18:00] It also shows me that Philadelphia's distribution [18:02] center has become a bottleneck with a really high return rate. [18:05] And finally, it shows me the breakdown based off [18:08] of my traffic source and its return rate. [18:11] So overall pretty great report. [18:14] As you see, our agent capabilities [18:16] have helped all the way from web querying to the web reporting. [18:22] With that, let me invite mani, who's [18:24] going to talk about our capabilities [18:26] that our developer facing. [18:29] Thank you. [18:34] Thank you very much, Ganesh. [18:36] Now let's switch gears and talk about how do you [18:38] build your own data agents. [18:41] Ganesh actually showcased some of the awesome capabilities [18:44] of conversational analytics. [18:46] So all the conversational analytics capabilities [18:49] that he showed was available in Google Cloud [18:51] Console in our one piece service or our first party service. [18:55] One of the constant feedback that we've [18:57] heard from our customers is, I really [18:59] like conversational analytics, but I [19:00] want to be able to take that and embed it in my own applications, [19:04] whether it's my Chat application or my custom app. [19:08] And today, to answer that, we are [19:10] very thrilled to announce the general availability [19:12] of conversational Analytics API. [19:14] Through this, customers will be able to take and embed [19:18] conversational analytics capabilities in their own custom [19:21] applications. [19:22] And CA API has got all the power and capability [19:27] and all the governance that is put [19:28] in place to ensure that your users get access [19:31] to the right data, and it to all the data sources [19:35] that Ganesh alluded to earlier, such as BigQuery, Looker, [19:39] AlloyDB and other databases. [19:43] The other thing that we also wanted to talk about [19:45] was the general availability of BigQuery SDK. [19:49] So this is for customers that are [19:51] using Google SDK to basically author their own agents. [19:55] Now with BigQuery SDK integration, [19:57] customers can seamlessly access built in BigQuery toolsets [20:01] directly within the SDK framework. [20:05] The other thing that we are also generally making generally [20:08] availability is the BigQuery MPP server. [20:11] This is a fully managed remote MPP [20:14] server that gives complete access to BigQuery, [20:17] and it eliminates the need for you [20:19] to create custom glues or API code to connect to BigQuery. [20:22] So this is also generally available, [20:24] and it's a fully managed service for customers [20:27] that want to use open source frameworks [20:30] or open source MCP tools to actually manage your own MCP [20:33] servers. [20:34] We are also thrilled to announce the general availability of MCP [20:37] toolbox, which is going to give you all the capabilities but you [20:41] as a user, can actually manage this environment. [20:46] For context, BigQuery MCP is one of the fastest growing area [20:51] in our portfolio. [20:53] As you can see from this particular data set, [20:55] we have the number of jobs that has been [20:58] processed to through BigQuery. [20:59] MCP has grown 20 year over year and it's [21:03] totaling millions of jobs. [21:04] So it's quite in the last six months, by the way, [21:07] and it's actually one of the fastest growing areas for us. [21:11] Now that you've seen all the capabilities [21:13] in terms of the agents that we've launched, [21:15] how you can customize and embed the agent. [21:18] One of the constant feedback that we hear from our customers [21:21] is, I want to understand how my agent is [21:23] performing in real time. [21:25] So they want to be able to have agent analytics to understand [21:28] how your agent is performing. [21:30] What is the latency of the agent, how many tokens the agent [21:33] is using, what is the customer sentiment, and so on and so [21:35] forth. [21:36] To address this, we are thrilled to announce [21:39] the general availability of BigQuery agent analytics plugin [21:42] with a single line of code. [21:44] If you are using frameworks like SDK and graph, [21:46] you can actually instrument and ensure [21:49] that you are able to get complete access on how [21:52] your agent is performing in real time [21:54] and all the data is going to be persisted in BigQuery. [21:57] So you as a user can go ahead and query the agent [22:00] to understand or query the BigQuery [22:02] table to understand how your agent is actually performing [22:05] and in fact a tip. [22:07] You can also bootstrap conversational analytics [22:10] on top of it and be able to interact with it [22:12] in a natural language way. [22:17] Now let's switch gears and talk about the Google Cloud data [22:20] Agent Kit. [22:21] So one of the constant feedback that we've [22:23] heard from our developers, the data practitioners, the data [22:26] engineers, and the data scientists [22:28] is that they use different coding environments. [22:31] Some of them are using Code, some of them [22:34] are using terminal based coding agents such as Cloud, code, [22:37] Gemini, CLI, Codex. [22:39] And today, in order for them to use these agents, [22:42] they have to actually customize their own skills, [22:45] their own tools and such. [22:46] They asked us for a capability that's [22:48] going to simplify this workflow. [22:50] So yesterday we announced the preview [22:52] of Google Cloud data Agent Kit. [22:55] And that's essentially comes in two flavors. [22:58] One is a VS Code extension. [23:00] So you can go ahead and install the data Agent [23:02] Kit in from the VS Code marketplace [23:05] and be able to embed it directly in your VS Code. [23:08] And the other thing is, it also comes [23:09] with built in plugins for Gemini CLI, Cloud Code and Codex. [23:14] And for context, the skills and the MCP tools [23:16] are open source so you can take it. [23:19] You can actually customize it based on your environment, [23:21] your use cases and such. [23:23] And we are very confident that this [23:25] is going to rapidly accelerate developer velocity. [23:29] With that, let me show you a quick demo of the data Agent [23:32] Kit. [23:35] So what you're going to see is the onboarding experience [23:39] of the data Agent Kit. [23:40] So here once you install the extension [23:44] you can go through the Configure section. [23:46] And here we will give you all the MCP servers that you [23:49] can actually configure. [23:51] And this is something that's available [23:52] for you right as a part of a simplified experience. [23:56] Now once you go ahead and configure [23:58] the MCP service of your choice. [24:00] You can go ahead and get started. [24:06] There you go getting started. [24:08] And here you will be able to configure [24:10] your complete environment in terms of which data set you [24:13] want to select and all that. [24:15] And it also comes with the predefined skills [24:16] that I just explained to you before. [24:18] Now we can take these skills. [24:20] You can understand what these skills are. [24:22] And you will be able to also go ahead and edit and customize [24:25] the skill if you wish to do so. [24:30] The Agent Kit has also got exploration capabilities, [24:35] so here it's got this catalog viewer [24:37] where you'll be able to see all the data sets that's [24:40] in your Lake house, your BigQuery, your AlloyDB [24:43] and Spanner. [24:44] You also will be able to actually [24:46] quickly search for the tables that you want, [24:48] and be able to interact with it. [24:50] In the next session, I'll show you how to chat with the data. [24:53] It's got a catalog viewer where you'll be [24:55] able to see your GCS objects. [24:58] Here you'll also be able to see your BigQuery jobs. [25:01] It's also going to show you your saved queries [25:03] and so on and so forth. [25:05] So in a nutshell, it's going to give you a very simple interface [25:08] through which you will be able to have quick access to the data [25:11] without having to write multiple SQLs and such. [25:14] It's called a spark viewer. [25:16] You will also be able to see your data engineering pipelines [25:19] right within the editor itself. [25:24] And finally, you have the setup page [25:26] where you can come in, configure what you need to do, [25:29] or if you wish to learn about things you can do. [25:31] So now once you have installed and explored. [25:36] So this is the capability within the Agent Kit [25:38] where you can simply chat with the data. [25:40] You do not have to write any kind of SQL [25:42] so you can simply in a natural language [25:44] way you can say, tell me more about this data set, or tell me, [25:49] the Q4 revenue and so on and so forth, [25:51] and the product is going to automatically give you [25:53] the results. [25:54] So as developers, you do not need to write SQL. [25:58] You can just interface with it. [25:59] And this is, by the way, powered by conversational analytics. [26:02] So we are bringing conversational analytics [26:04] capability right to the developers [26:07] within VS Code itself. [26:08] It not only produces the results, [26:10] it produces the charts and so on and so forth. [26:13] Once you've explored the data, you [26:14] can simply interact with the agent through a prompt. [26:17] Here I'm going to go ahead and do a fraud detection use case. [26:22] So it's a bunch of code that you're going to see scrolling. [26:25] But in a nutshell, what you're going to see is the agent [26:27] is going to do a planning for you based on the skills [26:30] that we've actually set up. [26:32] It is going to go ahead and generate the data pipeline [26:35] code for you. [26:36] It's going to do feature engineering for you. [26:37] It's going to clean up the data, and it's [26:39] going to create the pipeline right from within the interface [26:42] itself. [26:43] So you as a developer do not have to worry about it. [26:46] You can simply give the instructions. [26:48] It not only gives you the pipeline, [26:50] it also gives you the DAG of how the pipeline is being assembled. [26:54] So as you can see, it is going to vastly simplify [26:57] the workflows for developers. [27:00] Then you can even run the pipeline from here [27:02] to test whether the pipeline is working. [27:03] So you do not have to go to 10 different interfaces [27:06] to do this work. [27:07] It is automatically done from within this interface itself. [27:11] Now, once you've created this pipeline, the next step that you [27:15] can actually do this is still the pipeline stuff going on, [27:19] which is like once the pipeline is run, [27:21] you will be able to see whether the pipeline is run from it [27:24] from within the interface. [27:25] So as a developer, it is going to swiftly accelerate [27:27] your productivity. [27:29] Once your pipeline is done, you can actually do model creation. [27:33] You can do comparison of models from within it. [27:35] So with a simple single prompt you [27:37] can say, OK, compare two different models [27:39] and give me the output for me from within the interface [27:43] itself. [27:43] So the Agent Kit is going to again use [27:45] the skills behind the scenes, which is the data science [27:48] skills that we've shipped. [27:49] And it's going to go ahead and create all the scaffolding [27:51] code for you. [27:53] It not only creates the scaffolding code, [27:54] it's also going to produce the comparison within the Ide [27:59] itself. [28:00] So to summarize, we've actually vastly accelerated the developer [28:05] productivity using data Agent Kit. [28:07] And we can't wait to see what you [28:08] are going to build from this. [28:10] With this I'm going to actually bring [28:12] in Mauro, who's going to talk about his success story [28:15] in terms of leveraging data agents. [28:17] Mauro don't you love it when you're the last one to go. [28:29] My team is here, by the way. [28:30] I want that. [28:32] So who am I. What am I doing here. [28:36] Not telling you more about the product. [28:37] I'm telling you more about what we've done. [28:39] But before I tell you about that, [28:40] tell you about me really quickly. [28:43] Why this is interesting. [28:44] Because I'm not going to do the PR piece. [28:46] What I'm going to tell you is really what we have implemented [28:49] and what's working. [28:50] Where do we have some where do we [28:53] have some challenges in the journey [28:55] and how do we work around those. [28:58] So myself, I'm originally from Mexico, but I'm based in the UK. [29:01] I lived there for many years. [29:03] I've done tech for over 25 years. [29:06] So why telco? [29:07] That's been my last three years, really complete [29:10] different industry. [29:13] And it's really interesting because when you hear telco, [29:16] most of the times you hear churn and I'll tell you [29:20] more about that in a bit. [29:21] I'm a martial artist as well. [29:23] I've done martial arts for over 30 years. [29:25] I love it, I'm passionate about it. [29:27] So if you want to talk about that as well, [29:29] happy to talk to you about that. [29:31] More importantly, I'm a father and a husband. [29:34] I think I want to identify myself [29:35] as that family first and everything I do. [29:39] Basically, that's what guides me. [29:41] So let's go a bit more into VMO two and Virgin Media O2. [29:47] We tend to just minimize that VM. [29:51] One of the two largest providers of telco in the UK. [30:00] We combine mobile and broadband together and we are, [30:05] I'll say, the second largest in the UK. [30:07] We have about 45 million connections. [30:10] This is between broadband, mobile, TV, you name it. [30:13] So this should give you a bit of a scale [30:16] piece of what we're trying to achieve with everything you saw. [30:20] On top of that, of course, we are working towards a net zero [30:23] gain. [30:24] We also are feeling about 99% on mobile of the entire population [30:32] in the UK. [30:33] And then we're planning to be fiber first [30:40] by 2028, 80% of the UK. [30:43] One of the big changes and what attracted me to go into telco [30:46] is telco was a lot about building infrastructure. [30:51] Put the cables, put the antennas. [30:54] But how everything has shifted actually [30:58] needs a different approach. [30:59] It needs tech to make it happen. [31:02] So if I move on, what am I doing or what are we doing there. [31:07] That is not just me, it's my entire team. [31:10] One thing that we wanted to get right [31:12] is if you first start with the infrastructure. [31:15] Yes, you have a lot of power. [31:16] You have all those connections. [31:18] Think about your case. [31:19] When you buy a house rent a house rent an apartment, [31:21] you come in Connect yourself, and you [31:23] pay to the mobile provider or the broadband provider [31:26] that you have around you. [31:28] That was our model. [31:29] And not because we didn't want to do better. [31:32] That's how we used to operate. [31:35] How do you shift that big, heavy machine [31:38] into a more Agile machine. [31:41] Thinking about what the customer needs [31:43] is where data and I kind of came into place. [31:46] So we've been in the journey for the last two years and a half, [31:49] and the real focus has been on first centralizing the data. [31:54] That's part of another part of our teams [31:56] that we decided, let's move everything to GCP. [32:00] And we're 16,000 people, 16,000 employees, [32:03] different organizations, different systems, [32:06] moving everything to GCP. [32:08] That was the first part of the journey. [32:11] Then from there, we decided to build everything on BigQuery. [32:14] Why is this relevant for anyone here. [32:16] Because the reality is again, yes, you [32:17] saw what the products can do. [32:20] Again rubbish in, rubbish out. [32:22] So what do we do then. [32:26] Our strategy has been BigQuery is the single source of truth [32:30] and we build business layers. [32:31] Meaning we understand the business, [32:34] we bring the business in the journey. [32:36] And everything we built in BigQuery [32:38] is single source of truth for that business domain. [32:41] How do we connect. [32:42] All of that is a big part of the challenge. [32:44] From there, the challenge was how can people [32:47] access that data to make decisions to improve customer [32:50] experience and bring better revenue for our business. [32:54] That's quite a journey, right. [32:55] So we at that point you go for self-service. [32:59] Think about it. [32:59] Dashboards right. [33:00] So you have all your data in BigQuery. [33:02] All good. [33:03] How do you want people to access that data. [33:05] You put a dashboard. [33:06] All good. [33:07] If you're good at building that and you understand the business [33:10] you got it right. [33:11] But how long does it take to democratize all that data [33:15] if you're building assets and objects. [33:18] It's quite a journey, right. [33:19] And you keep building and building and building [33:21] and discoverability becomes a bit of a trick. [33:24] So that's when AI and AI comes to play. [33:28] If any of you work with data, you [33:30] might have seen your backlogs. [33:33] They're never ending, right. [33:34] You always keep piping stuff in. [33:39] So conversational analytics and why we started the journey [33:44] last year with Google. [33:46] They were early in their journey. [33:48] We partnered with them on that. [33:50] Number one thing we wanted to do this is a great opportunity [33:53] to get data to the hands of our customers or stakeholders ASAP. [33:58] Why Because we build the foundations. [34:00] So how do we do that. [34:01] How do we make it easy for them to have [34:03] that natural conversation with data so they can [34:07] get the results that they need. [34:09] Then we wanted to make sure that people [34:11] were able to find what they were looking for without having [34:15] to talk to us. [34:16] Yes, we were great. [34:18] I love my team. [34:18] I love people talking to my team, [34:20] but that's not the best use of our time. [34:23] We're about the data size. [34:25] We're a team of combining all of us. [34:28] I think we might be about 800 people in data. [34:33] It's a lot of people. [34:36] So how do we free up the time for everyone. [34:38] And they're not only on my team. [34:40] How do we free time of these people [34:42] to make sure they're investing time where it's needed. [34:44] Democratize access. [34:45] Make it discoverable. [34:47] And then we wanted people to be curious. [34:51] Don't you get normally people we're talking about great tech, [34:54] great tech without deriving value means nothing. [34:57] In the end, you'll have fun with the toy. [35:00] But if it's not really driving the results, no one cares. [35:03] So we want people to be curious. [35:05] We want people to be able not to just ask one question what's [35:09] the revenue today. [35:10] Yeah that's fine. [35:12] I want you to explore. [35:13] I want you to get further, because that's the only way [35:15] we can change the game, right. [35:16] People don't think the same way today [35:18] as they were thinking yesterday. [35:23] So here it goes. [35:24] The interesting piece. [35:25] This was all my team. [35:28] We built something on top of CDK. [35:32] We named it parity. [35:33] There was a debate on the name. [35:35] So if you don't like it give us feedback anyway. [35:37] But we've been calling that for the last six months. [35:39] So parity. [35:42] It's powered by Gemini enterprise in the end. [35:46] This is part of the architecture of what we do. [35:49] So business layers became that foundational layer. [35:53] What's the tricky piece with data agents. [35:56] You're talking about answering business questions that [36:01] are critical to make decisions. [36:03] A simple example, what's the revenue. [36:06] I go back to the revenue point. [36:08] If you think about that simple question, [36:10] if the agent gives you two or three different answers [36:14] every time you query it, that's a massive issue. [36:18] So having that foundational layer that we can certify [36:21] and we can trust was key. [36:22] So we pulled that from BigQuery. [36:25] We configure all of that into the agent. [36:27] And then we call our Analytics API. [36:31] We're able to save all of these into favorites. [36:33] You'll see a demo in a bit. [36:34] So that first layer that you see on top [36:36] is all reliant on those foundational layers in BigQuery. [36:39] Then my team did something amazing. [36:41] And I think this is something that if you [36:43] are using these frameworks and these tools, [36:46] you might benefit from, they built something on top that. [36:50] When we started the journey, the exploration [36:54] was about three months. [36:55] From where we were POC get something live. [37:00] Once we were in production and they'll kill me for saying this. [37:03] But now I saw we can create an agent in minutes. [37:08] Now if someone from VMO two listening to this in days. [37:12] Yeah, but minutes. [37:15] I'm not. [37:15] I'm not the PR guy. [37:16] It's minutes. [37:20] The power of that is crazy. [37:23] Now, if you have the right context, [37:24] you've done the right thing. [37:26] This works. [37:27] If you haven't, guess what. [37:28] You're going to hit privacy. [37:29] You're going to hit governance. [37:31] You're going to see hit legal. [37:32] All of those things. [37:33] So really important to invest in that piece. [37:35] And then the bottom layer right is [37:38] how we're managing the feedback to keep improving [37:40] what we're doing right. [37:41] How do we build that consistency in an ongoing basis [37:44] to make sure that this works. [37:47] So from prototype to empowerment, three simple steps. [37:51] Rapid prototyping. [37:52] We're always afraid of how this is going to work. [37:55] If you're on the tech space, if you're on the engineering space, [37:58] getting the business buy in was key. [38:02] How do you build the approvals quickly. [38:05] We were really aggressive with our procurement team. [38:07] They're really friendly and to us, [38:10] but we had to be quite aggressive there. [38:13] But also we had to prove that we were going to keep this safe. [38:18] So one investment we did that might be helpful for you or not, [38:21] is we committed to train every person that [38:24] was going to use conversational analytics to make [38:27] sure they use it safely. [38:30] So everyone that goes into one of the licenses that we have, [38:34] they get onboarded and they have to get trained. [38:36] Otherwise, they lose access to the tool. [38:40] And then scaling right now. [38:43] So at the moment, we put this together, [38:45] we were talking about hundreds plus employees. [38:47] I'll give you the numbers in a bit. [38:49] So let me show you a quick demo. [38:50] I'm going to pass it over. [38:52] So I'll do the best I can on that one. [38:54] Think about before I started, think about one [38:56] of the typical use cases of telco. [38:58] I'm sure many of you are there from your mobile service [39:01] or your broadband service. [39:03] You've been on the phone talking to someone in customer contact [39:07] to say, hey, I have an issue. [39:08] Can you help me. [39:10] That's the most engagement that we have. [39:13] A lot of engagements on that space, [39:14] improving how we interact. [39:16] Either we embed AI or not is really important. [39:19] So we created an agent as a disclaimer. [39:22] The data has been modified. [39:23] This is an actual setup. [39:26] This an actual agent. [39:28] The numbers you're going to see are not our numbers. [39:30] So just don't go with that. [39:32] So in our enterprise we start from there [39:36] going to our enterprise. [39:38] Our users can go and explore the different agents that we have. [39:41] In this case, we're going to show the customer interaction [39:44] agent. [39:44] One of typical questions we want to go [39:46] for is, hey, how many calls have we had in the last few months. [39:53] Now that's a simple question, but think about the complexity [39:56] of that question. [39:59] If you're on the business side, calls can mean so many things. [40:03] So we ask that this is going back [40:05] to BigQuery to one of our business layers, [40:07] querying that data. [40:09] And then it's coming back with a result. [40:10] Now whilst it's thinking and coming back with a result, [40:13] it's saying actually what do you mean by coal. [40:18] If this was one of our data analysts or one [40:21] like it will have said 10, but actually this [40:25] is asking for more context to make sure we get it right. [40:28] So we go back, we provide the context that it needs and then [40:31] goes back and queries the data. [40:32] Now normally if you're in a dashboard, [40:34] you will typically stop there or try [40:37] to drill down or ask where do I find the next question. [40:40] But then what if I want to drill down and split this by channel. [40:45] What if I want to understand how those calls happen [40:48] and what's the pattern. [40:49] So I can actually go and change behavior, deploy better, [40:53] like deploy the right people at the right time or the resources [40:57] that I have. [40:57] So I asked, give me this by channel. [41:01] But guess what. [41:02] This thing is smart enough to tell me. [41:04] Well, you already asked me for this particular piece. [41:07] So let's go and show you. [41:13] Call and the two channels that we're displaying here, [41:16] calling messenger. [41:17] So it's actually challenging me back to say, [41:20] is this what you mean all the time to make sure [41:24] that the user have that control and that input back into it. [41:28] Now, one of the other key pieces that we have is [41:32] tell me how you did it. [41:33] How many times have has that happened. [41:35] Someone opens a dashboard. [41:37] You don't know how they came up with that conclusion. [41:40] Big part of the advantage of these agents [41:42] as well is we are able to go back and explain it. [41:45] What if I want to see it differently. [41:47] What if I want to trigger a different mindset. [41:49] Show me a different pattern. [41:50] Line graph would be a good one. [41:52] That's what you saw there. [41:54] Another key feature that we have is, [41:57] from a day to day perspective. [42:00] You go through queries constantly. [42:03] A feature that we leverage a lot is favorites. [42:06] So save this thing into my favorites. [42:09] And within those favorites, I can always go back [42:11] and create those favorites. [42:12] Or I can update that data whenever I want. [42:14] So the data can be refreshed, can be updated. [42:16] So how does this normally happen. [42:19] You'll have someone business user data user coming back [42:23] to member of my team. [42:25] Hey you know what I want a data agent that helps me do this. [42:29] This is what my team built on top of the APIs [42:33] that we have available. [42:34] So creating an agent is as simple [42:37] as you're going to see right now, right. [42:41] Use a name. [42:43] Please use good names. [42:44] Hopefully Barrett is a good one. [42:47] Pick up the tables that you want to enable for people to go [42:50] and see that they can have access to. [42:52] Then we created an agent on the back that actually goes and gets [42:56] more context builds more details that make that more accurate. [43:01] So this is what you're seeing at the moment right. [43:03] So the agent's going back doing some extra work, [43:06] filling up all that YAML file that you [43:08] need to send it to the API to do the right thing. [43:12] And then it comes back with the answers part of those answers. [43:15] It will give you the opportunity to fill up [43:17] some of the properties that you need for that agent function. [43:22] But the other more important thing that I find [43:24] is this part where it goes into the details on field by field. [43:31] This is where if you have the right context, then [43:33] this works well. [43:34] Where you don't have the right context then [43:37] this completely fails. [43:38] So having that agent automating that on the back end [43:41] makes this agent. [43:42] So accurate. [43:43] How much it has helped us. [43:45] We were not doing this before. [43:46] This is something we implemented later. [43:49] We have the option to modify it live or accept it as is. [43:54] The other next feature that we love is golden quarries. [43:57] So golden quarries. [43:58] What they do is when you know the answer to a question what [44:03] query you're expecting to get. [44:05] So we tune up the model by making sure [44:07] that when it goes and someone asks that question, [44:09] it's going for that query. [44:11] We can test it in this application that we created [44:15] and see that it's actually returned the right thing. [44:18] We accept it goes into the properties file that goes [44:22] that gets sent to the agent. [44:26] Simple thing. [44:27] Again, it shows you that you can deploy the agent anywhere [44:29] in the same test environment. [44:31] We can deploy the agent, showcase what the agent can do [44:34] and that's it. [44:36] Then at the end, we go back to the customer [44:38] and say, hey what agent created. [44:40] You can go and access it on Gemini app enterprise. [44:42] Of course, we have to provide the right access. [44:45] And there you go. [44:47] An agent in ECC. [44:49] Yeah so. [44:54] I'm on time, so I'm just going to quickly tell you [44:57] this is the business. [44:58] I want to have picked up this one. [44:59] I have a better one. [45:00] I got an email from one of the users that said, [45:04] people that have been in telco for 20 plus years as well, [45:06] right. [45:07] This is the best tool I've ever seen built in BMO two. [45:11] I promise you, this is not a lie. [45:13] I got an email like that. [45:14] So this is amazing. [45:16] This has a lot of impact on your users. [45:18] We're just starting. [45:19] We have 5,000 licenses that we want to complete by the end [45:21] of the year. [45:22] We're on 700 today 750, but 12,000 answers provided. [45:26] Thank you. [45:30] Moving to Ganesh to close. [45:33] Thank you Ganesh. [45:33] Thank you. [45:34] Thank you all. [45:35] So you all saw the transformative capabilities [45:38] of our agents. [45:40] So we also have more sessions in case [45:42] if you want to do the deep dive on our agents and capabilities. [45:45] Thank you all.