LXILC Launches Conclave Multi-AI Engine for Wargaming, Trading

Conclave AI Engine: From War Gaming to Trading & Complex Questions With LXILC — Ai4

Ai4

Structuring transparent, multi-round debates across competing frontier models generates institutional-grade strategic simulations and trading analysis that single-agent systems cannot match.

AI-assisted summary of Conclave AI Engine: From War Gaming to Trading & Complex Questions With LXILC — original source. Check the original for full context.

Source published:

Detailed notes

## Background Published by Ai4 on YouTube on August 24, 2025, this presentation features Alexandre Mogilamy, founder of Legacy eXperience Initiative Limited Company (LXILC). The video showcases the Conclave AI Engine, a multi-agent reasoning platform designed to orchestrate frontier AI models for real-time geopolitical wargaming, hedge-fund-level financial trading, and multi-perspective question answering.

## TL;DR Alexandre Mogilamy introduces the Conclave AI Engine, an architecture that coordinates seven frontier models (including Claude, Gemini, GPT, Grok, Mistral, Perplexity, and Cohere) across structured evaluation rounds to eliminate single-agent blind spots. The platform powers strategic wargaming simulations for geopolitical conflicts and corporate stress-testing by assigning discrete operational briefings and competing roles to individual models. It also drives Conclave Trading, which processes millions of tokens of financial feeds to provide rapid, multi-horizon technical and derivative market insights. Mogilamy concludes by sharing that he built the entire platform in four and a half months while working full-time as a bartender at the Bellagio in Las Vegas.

## Atmosphere & style The presentation has the earnest, fast-paced atmosphere of an ambitious tech demo delivered by a determined outsider. Mogilamy balances complex architectural breakdowns with accessible, real-world analogies, famously comparing prompt engineering parameters to the exact measurements of craft bartending. The talk builds an inspiring emotional arc, shifting from high-stakes military and financial simulations to a humble, empowering proof that deep technical innovation can emerge outside traditional tech pedigrees.

## Key takeaways - [00:04](https://www.youtube.com/watch?v=p47a2t060a8&t=4s) Conclave replaces opaque, single-AI responses with a transparent multi-agent engine orchestrating seven frontier models to evaluate, challenge, and refine complex queries. - [00:04](https://www.youtube.com/watch?v=p47a2t060a8&t=4s) The multi-round deliberation pipeline executes in 39 seconds, producing balanced answers alongside 'Black Swan' and disruptive critiques with a complete lineage audit trail. - [02:23](https://www.youtube.com/watch?v=p47a2t060a8&t=143s) Conclave War Gaming simulates geopolitical crises by assigning discrete AI agents to role-play competing actors such as the Pentagon, Congress, foreign governments, and financial markets. - [02:23](https://www.youtube.com/watch?v=p47a2t060a8&t=143s) Each simulated agent receives exhaustive briefing parameters, including mission profiles, strategic objectives, alliance protocols, and explicit rules of engagement. - [02:23](https://www.youtube.com/watch?v=p47a2t060a8&t=143s) The enterprise wargaming framework was applied to the pharmaceutical sector to model competitive rival moves, regulatory approval boards, and media coverage dynamics. - [02:23](https://www.youtube.com/watch?v=p47a2t060a8&t=143s) Rigorous parameter specification and prompt architecture enable the engine to deliver McKinsey-grade strategic scenarios in roughly four and a half minutes. - [05:40](https://www.youtube.com/watch?v=p47a2t060a8&t=340s) Conclave Trading ingests up to six million tokens of market data, deploying multiple agents to deliver hedge-fund-level technical, fundamental, and derivative analysis. - [05:40](https://www.youtube.com/watch?v=p47a2t060a8&t=340s) Future trading iterations are designed to integrate alternative data streams, including satellite supply chain imagery and sentiment and body language tracking from corporate earnings calls. - [08:18](https://www.youtube.com/watch?v=p47a2t060a8&t=498s) Future planned verticals include algorithmic sports betting, courtroom legal strategy simulation, medical research, and multi-agent personal conversational cohorts. - [08:18](https://www.youtube.com/watch?v=p47a2t060a8&t=498s) Founder Alexandre Mogilamy built the entire Conclave platform in four and a half months while working full-time as a bartender at the Bellagio in Las Vegas.

## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=p47a2t060a8&t=4s) Introducing Conclave: Beyond Single-AI Responses

Alexandre Mogilamy introduces Conclave, an engine designed to coordinate multiple frontier artificial intelligence models to solve complex reasoning problems. While existing platforms like ChatGPT operate on a simplistic "one question, one AI, one answer" paradigm, and recent iterations of Gemini or Grok route queries into opaque multi-agent pipelines where "nobody know where the answer go or is create," Conclave emphasizes total transparency and multi-perspective dialectic. When a user submits a question to the Conclave engine, seven distinct models analyze, challenge, and refine the query across structured evaluation rounds.

```mermaid flowchart TD accTitle: Conclave Multi-Agent Question Engine Architecture subgraph S1["Input"] Q["User Query"] end

subgraph S2["Round 1: Initial Drafting"] direction TB C1["Cohere: Baseline Formulation"] M1["Mistral: Alternative Perspective"] G1["GPT: Black Swan / Futuristic"] Gr1["Grok: Subversive / Disruptive"] end

subgraph S3["Round 2: Evaluation & Refinement"] direction TB Cl2["Claude: Improves & Ameliorates Cohere"] P2["Perplexity: Improves Mistral Alternative"] end

subgraph S4["Output Synthesis (39s Total)"] direction TB Gem["Gemini: Synthesizes Final Balanced Views"] Out["Full Audit Trail & Multi-Perspective Deliverables"] end

Q --> C1 & M1 & G1 & Gr1 C1 --> Cl2 M1 --> P2 Cl2 & P2 & G1 & Gr1 --> Gem Gem --> Out ```

The refinement cycle functions in distinct phases. In the first round, models generate an initial balanced proposition alongside alternative answers, while two dedicated agents produce divergent angles—specifically a "Black Swan" futuristic forecast and a subversive, disruptive critique. In the second round, the initial drafts pass to competing models for rigorous critique and amelioration:

- Gemini synthesizes the primary answer, which was initially drafted by Cohere in round one and refined by Claude in round two. - A secondary alternative answer originates with Mistral in round one and undergoes optimization by Perplexity in round two. - GPT provides the unconventional "Black Swan" perspective, while Grok delivers the counter-narrative, disruptive analysis. - The entire deliberative pipeline executes in 39 seconds, granting users full visibility into the lineage and evolution of every generated perspective rather than an unverified single output.

---

[02:23](https://www.youtube.com/watch?v=p47a2t060a8&t=143s) Conclave War Gaming: Strategic Simulations and Enterprise Scenarios

Mogilamy explains how the Conclave engine extends naturally into geopolitical and enterprise wargaming by assigning distinct AI models to role-play competing factions within complex geopolitical crises. Presenting a live simulation of a South China Sea maritime escalation, Conclave designated discrete models to represent the Pentagon, China, Japan, the United States Congress, and Wall Street. To orchestrate authentic interplay, the engine feeds each agent an extensive briefing prompt containing situational context, overarching strategic objectives, key geopolitical parameters, mission profiles, alliance coordination protocols, and explicit rules of engagement.

The system synthesizes the inter-agent clash into high-level headline decisions for each faction, accompanied by an initial 380-character executive summary and an exhaustive 3,874-character breakdown of the geopolitical reverberations. Beyond military simulations, Mogilamy was approached by one of the largest pharmaceutical companies in the world seeking to stress-test their corporate strategy. Conclave mapped out their competitive landscape by simulating competing pharmaceutical rivals, regulatory review boards evaluating medication approvals, and media dynamics—modeling specifically how major outlets like Bloomberg report on corporate milestones and whether secondary journalists amplify stories when primary outlets face conflicts of interest.

Drawing on his personal experience as a bartender, Mogilamy illustrates that the true bottleneck in artificial intelligence is precise parameter specification. When peers doubt AI's capacity to design craft cocktails, he asks whether they specified the exact acidity of the lemon juice (such as 0.7), the specific gravity of the simple syrup, glassware selection, ice dilution rates, and mixing technique. "AI today they can do everything," Mogilamy notes; "we just have to know to talk to them to ask the perfect question with all the data." Within Conclave, the engine acts as the intermediary bridging complex human objectives with rigorous, multi-agent prompt construction, delivering comprehensive enterprise scenarios in four and a half minutes that rival deliverables from traditional firms like McKinsey.

---

[05:40](https://www.youtube.com/watch?v=p47a2t060a8&t=340s) Conclave Trading: Hedge-Fund-Level Multi-Agent Market Analysis

Conclave Trading deploys multi-agent reasoning to evaluate financial markets, seeking to provide individual investors with the analytical rigor typically restricted to elite institutional hedge funds. Ingesting high-volume market feeds—handling as many as six million tokens of end-of-day and technical data—the engine splits the analysis across multiple specialized models. These agents parse technical chart formations, fundamental metrics, and indicators to synthesize unified short-term, mid-term, and long-term trading horizons.

The trading module processes standard financial inquiries in 1 minute and 10 seconds, with deep-dive strategic assessments completing within two to three and a half minutes. Future iterations of the model are designed to integrate non-traditional alternative data sources, such as AI-driven sentiment and body language analysis from corporate earnings calls, as well as satellite imagery tracking manufacturing and supply chain output. The system is designed to support direct derivative inputs:

- Stocks and complex equity options tracking strike prices, expirations, and historical volatility. - Flexible analysis accommodating pre-trade planning or real-time portfolio management on already-open positions. - Institutional hedge-fund-grade risk evaluation that replaces gut feelings and speculative chart-gazing with multi-perspective probabilistic modeling.

---

[08:18](https://www.youtube.com/watch?v=p47a2t060a8&t=498s) Expanding the Frontier: Sports Betting, Legal Strategy, and Alexandre's Journey

Beyond questions, wargaming, and trading, Mogilamy outlines broad future verticals for the Conclave platform. These include algorithmic sports betting models analyzing complex fight odds; multi-agent legal companions that simulate courtroom proceedings by modeling judge dispositions, prosecutorial maneuvers, and defense strategies; and specialized engines configured for cryptocurrency markets, medical diagnostics, and scientific research. He also envisions conversational companions where a user talks to a cohort of five differentiated AIs at once—a supportive friend, a trusted confidant, a practical advisor, and a critical devil's advocate challenging assumptions.

Closing his presentation, Mogilamy shares the personal backstory behind Conclave's development. Operating entirely outside the traditional technology sector, he conceived and engineered the platform from scratch over just four and a half months while working a full-time job as a bartender at the Bellagio in Las Vegas. Demonstrating that ambition and prompt architecture matter more than conventional tech pedigree, he directs attendees to test the free live engine via QR code or visit Booth 454, where dedicated tablets were staged for live user testing.

## Notable quotes > [00:04](https://www.youtube.com/watch?v=p47a2t060a8&t=4s) nobody know where the answer go or is create — Alexandre Mogilamy

> [02:23](https://www.youtube.com/watch?v=p47a2t060a8&t=143s) AI today they can do everything — Alexandre Mogilamy

> [02:23](https://www.youtube.com/watch?v=p47a2t060a8&t=143s) we just have to know to talk to them to ask the perfect question with all the data. — Alexandre Mogilamy

> What if you could simulate and manage the most critical geopolitical crises, test your strategic decisions, and anticipate the unpredictable—in real-time, powered by AI?

> Conclave War Gaming isn’t a simulation—it's strategic foresight powered by advanced, collaborative AI.

> Markets never sleep—neither does Conclave Trading.

> Complex questions deserve more than a single answer.

> Conclave isn’t just AI answering questions—it's intelligence debating, strategizing, and evolving.

## Mentioned Alexandre Mogilamy, Legacy eXperience Initiative Limited Company (LXILC), Conclave AI Engine, Conclave War Gaming, Conclave Trading, Ai4, ChatGPT, Gemini, Grok, Cohere, Mistral, GPT, Claude, Perplexity, Pentagon, United States Congress, Wall Street, Bloomberg, McKinsey, Bellagio

## Video description Alexandre Mogilamy, FOUNDER, Legacy eXperience Initiative Limited Company (LXILC)

Abstracts What if you could simulate and manage the most critical geopolitical crises, test your strategic decisions, and anticipate the unpredictable—in real-time, powered by AI?

1. Conclave War Gaming Imagine facing a sudden escalation in the South China Sea or managing a diplomatic standoff between global powers. Conclave War Gaming places multiple specialized AI agents into realistic geopolitical and military scenarios, each role-playing distinct actors (Pentagon, Congress, Foreign Governments, Market Analysts). Engage in immersive simulations, uncover hidden risks, and test strategic options under pressure. Conclave War Gaming isn’t a simulation—it's strategic foresight powered by advanced, collaborative AI.

2. Conclave Trading Markets never sleep—neither does Conclave Trading. Our advanced AI engine analyzes stocks, identifies signals, anticipates trends, and delivers sophisticated financial insights usually accessible only to top-tier hedge funds. It transforms massive financial datasets into actionable strategies by leveraging multi-agent reasoning, combining rigorous technical analysis with intelligent predictive modeling. With Conclave Trading, navigate uncertainty, anticipate market moves, and seize opportunities ahead of the curve.

3. The Conclave (Questions) Complex questions deserve more than a single answer. Conclave leverages multiple advanced AI agents—including GPT, Claude, Perplexity, Grok, and more—to debate, reason, challenge each other, and converge towards nuanced, insightful responses. Ask philosophical dilemmas, strategic business problems, or real-world questions, and watch the Conclave engine transform isolated answers into dynamic dialogues and thoughtful, collective insights.

Powered by the Conclave AI Engine At the heart of all these applications is the Conclave Engine: a modular, multi-agent reasoning platform designed for complexity and uncertainty. It enables curated scenarios, intricate role assignments, and advanced logic interactions—delivering insights far beyond single-agent solutions.

Conclave isn’t just AI answering questions—it's intelligence debating, strategizing, and evolving. Join us and experience the power of collaborative AI thinking.

[Source](https://www.youtube.com/watch?v=p47a2t060a8)

中文简报

LXILC推出用于兵棋推演与交易的Conclave多AI引擎 在竞争性前沿模型之间构建透明的多轮辩论，可生成单智能体系统无法比拟的机构级战略模拟与交易分析。 Alexandre Mogilamy 介绍了 Conclave AI 引擎，该架构通过结构化评估轮次协调七个前沿模型（包括 Claude、Gemini、GPT、Grok、Mistral、Perplexity 和 Cohere），以消除单智能体的盲区。该平台通过为各个模型分配独立的操作简报和竞争角色，为地缘政治冲突及企业压力测试提供战略兵棋推演模拟。它还驱动 Conclave Trading，后者处理数百万 token 的金融数据流，以提供快速、多周期的技术与衍生品市场洞察。Mogilamy 最后分享道，他在拉斯维加斯 Bellagio 担任全职调酒师期间，历时四个半月搭建了整个平台。 Conclave 推出透明的多智能体推理引擎，协调 7 个前沿 AI 模型在 39 秒内完成多轮辩论与修正，不仅输出均衡结论，还能呈现“黑天鹅”与颠覆性视角并附带完整的审计追踪。 (0:04) Conclave War Gaming 支持地缘政治与企业兵棋推演，通过为五角大楼、外国政府及市场等角色分配详细参数与交战规则，能在 4.5 分钟内生成媲美麦肯锡级别的战略模拟方案。 (2:23) 面向金融市场的 Conclave Trading 可吞吐高达 600 万 tokens 的行情数据，借助多智能体协作提供机构对冲基金级别的技术面与衍生品分析，未来计划整合卫星图像和高管肢体语言等另类数据。 (5:40) 该平台由拉斯维加斯 Bellagio 酒店调酒师 Alexandre Mogilamy 在业余时间历时 4 个半月自研开发，未来还计划拓展至法庭诉讼对抗、体育博彩与多角色 AI 伴侣等领域。 (8:18)

Reading summary

Alexandre Mogilamy presents Conclave, a multi-agent artificial intelligence platform designed to replace opaque single-model responses with transparent, multi-round deliberation among competing frontier models. He asserts that rigorous prompt parameterization enables coordinated AI architectures to conduct complex evaluations across geopolitical wargaming, corporate strategy, and institutional-grade financial analysis. Multi-Round Dialectical Architecture and Auditability (0:04) Mogilamy argues that existing conversational systems like ChatGPT rely on an oversimplified single-response paradigm, while platforms such as Gemini and Grok obscure how multi-agent answers are derived. Conclave attempts to counter this lack of transparency by routing user queries through seven distinct frontier models across structured drafting and refinement stages. The full deliberative workflow executes in 39 seconds, producing a verifiable audit trail that records the evolution of each argument rather than presenting an unsubstantiated single output. The pipeline begins with Cohere formulating a baseline answer and Mistral generating an alternative viewpoint, while GPT and Grok contribute divergent angles through a speculative 'Black Swan' forecast and a disruptive critique, respectively. In the second round, Claude refines Cohere's initial baseline while Perplexity optimizes Mistral's counter-proposal. Finally, Gemini synthesizes the refined positions into balanced conclusions alongside the contrarian tracks, allowing users to examine the lineage and interaction of each distinct model output. Strategic Simulation via Precise Context Parameterization (2:23) The engine extends dialectical modeling into wargaming by assigning competing frontier models to distinct geopolitical or corporate actors. In a demonstrated simulation of a South China Sea maritime escalation, discrete agents represented the Pentagon, China, Japan, the United States Congress, and Wall Street. Each agent was seeded with detailed situational context, strategic objectives, alliance coordination protocols, and rules of engagement, ultimately producing headline faction decisions, a 380-character executive overview, and a 3,874-character analysis of geopolitical outcomes. Conclave applied the same logic to corporate defense for a global pharmaceutical firm by modeling rival manufacturers, regulatory review boards, and reporting biases between primary news outlets and secondary journalists. Drawing on his experience as a bartender, Mogilamy claims that artificial intelligence capability is constrained primarily by how precisely parameters are defined. He illustrates this with cocktail design, where successful execution requires specifying lemon juice acidity levels such as 0.7, simple syrup specific gravity, ice dilution, and mixing mechanics. Mogilamy contends that when human operators supply complete data and explicit boundary conditions, multi-agent frameworks can generate enterprise-grade scenario analyses within four and a half minutes that rival the deliverables of traditional consultancy firms. Probabilistic Market Analysis and Financial Data Processing (5:40) Conclave Trading applies multi-model reasoning to financial markets with the goal of providing retail investors with institutional hedge-fund-level capabilities. The system digests up to six million tokens of end-of-day market feeds and technical indicators, distributing sub-tasks across specialized models to construct unified outlooks across short-, medium-, and long-term trading horizons. Standard market assessments complete in 70 seconds, whereas deeper strategic evaluations require between two and three and a half minutes. According to Mogilamy, the platform evaluates complex equity options by incorporating historical volatility, strike parameters, and expiration cycles for both pre-trade planning and existing portfolio management. He outlines plans to expand the trading engine's data ingestion beyond standard metrics to include alternative indicators, such as computer-vision tracking of satellite imagery over manufacturing facilities and automated sentiment and body-language analysis during corporate earnings presentations. Platform Origins and Proposed Domain Expansions (8:18) Beyond questions, wargaming, and equity trading, Mogilamy envisions deploying Conclave into algorithmic sports betting to evaluate fight odds, as well as legal strategy applications that simulate judicial dispositions, prosecutorial tactics, and defense arguments. Additional conceptual verticals include cryptocurrency analytics, medical diagnostic engines, and five-agent personal cohorts designed to simultaneously provide supportive, advisory, and adversarial devil's advocate feedback. Mogilamy developed the Conclave engine over four and a half months without a conventional computer science pedigree, building the software while working full-time as a bartender at the Bellagio in Las Vegas. He presents this background as evidence that granular prompt engineering and multi-agent coordination frameworks matter more than traditional institutional credentials in extracting advanced performance from frontier AI. Mogilamy positions Conclave as a demonstration that rigorous parameter specification can turn off-the-shelf frontier models into institutional-level analytical tools. However, several highlighted features—including satellite imagery ingestion, body-language analysis on earnings calls, and courtroom simulation modules—remain future conceptual expansions rather than verified, production-ready capabilities.

精炼阅读摘要

Alexandre Mogilamy 推出了名为 Conclave 的多模型协同人工智能引擎，该系统通过结构化的多轮辩论与全透明审计流程，旨在解决复杂问答、地缘战略推演和金融交易分析等领域的单模型偏差与不透明问题。 多智能体辩证架构与透明问答 (0:04) 传统人工智能平台（如 ChatGPT）主要依赖单模型生成回答，而 Gemini 或 Grok 的多代理架构在执行路径上往往缺乏透明度。Conclave 引擎则采用结构化的多轮评估机制，协调七个前沿 AI 模型对用户问题进行交叉审视与综合提炼。在第一轮起草中，Cohere 负责基准表述，Mistral 提供替代视角，GPT 负责提出具前瞻性的“黑天鹅”预测，Grok 则产生颠覆性批评。进入第二轮后，Claude 对 Cohere 的初始版本进行改进，Perplexity 则完善 Mistral 的方案。 在最终输出环节，Gemini 负责将经过两轮打磨的观点综合为平衡的主流方案，同时完整保留“黑天鹅”和颠覆性反叙事视角。整套推演流水线在 39 秒内完成，并向用户提供完整的生成脉络与审计线索。Mogilamy 指出，这种设计不仅杜绝了无法溯源的黑箱操作，还确保了输出结果的多样性与可靠性。 战略博弈模拟与企业竞争压力测试 (2:23) Conclave 引擎将其多模型辩证能力扩展至地缘政治及商业推演领域。以南海局势模拟为例，系统分别为五大独立模型分配了五角大楼、中国、日本、美国国会以及华尔街的角色，并向每个智能体注入涵盖战略目标、地缘政治边界、结盟协议及交战规则的详尽提示词。系统最终输出了 380 字符的执行摘要以及长达 3,874 字符的地缘冲击深度解析。 在商业应用方面，该系统为一家大型跨国制药企业开展了战略压力测试，模拟其主要竞争对手、药物审批监管机构以及包括彭博社在内的主流与次级媒体动态。Mogilamy 借调酒师的调饮经验解释称，提示词工程的精细参数设置（如特定酸度、糖浆比重和稀释率）是释放 AI 全面潜力的关键。Conclave 能够在四分半钟内生成媲美麦肯锡等传统咨询机构的企业战略方案。 面向金融交易的多代理投研与风控 (5:40) 针对金融市场分析，Conclave Trading 试图将对冲基金级别的投研能力普及至个人投资者。该模块能够处理高达 600 万 Token 的收盘和技术数据，通过专门的模型分工解析图表形态、基本面财务数据以及各类量化指标，最终生成涵盖短期、中期和长期的统一研判方案。常规财务查询可在 1 分 10 秒内完成，更深入的战略评估耗时约 2 至 3.5 分钟。 该模块支持股票及复杂权益期权输入，跟踪行权价、到期日与历史波动率，既适用于盘前规划，也适用于实时持仓管理。Mogilamy 规划中的未来版本还打算纳入非传统替代数据，例如财报电话会议中基于 AI 的情绪和肢体语言解析，以及监测制造业产能和供应链的卫星图像，以概率量化模型取代主观臆测。 长远垂直领域规划与草根开发历程 (8:18) 除了问答、战争推演与交易外，Mogilamy 设想将 Conclave 推广至体育博彩胜率计算、模拟法官与控辩双方互动的法庭策略助手，以及加密货币、医疗诊断与科研辅助工具。系统未来还规划了集成多重人设的情感伴侣功能，允许用户同时与支持者、密友、务实顾问和提出质疑的“唱反调者”五位智能体协同对话。 在个人背景方面，Mogilamy 并非传统科技从业人员，他是在拉斯维加斯百乐宫酒店全职从事调酒师工作的同时，历时四个半月独立从零设计并构建了 Conclave 平台。这一实践展示了合理的系统提示架构与调度能够跨越专业技术背景的壁垒，目前该平台已开放免费测试。 Conclave 展示了多模型协同推演在复杂场景下的应用潜力，但该系统的实际运行成本、对第三方基础模型接口稳定性的依赖程度，以及其推演预测在现实复杂环境中的有效性与合规风险，仍有待更长期的外部检验。

Detailed notes — zh

## Background 该视频由 YouTube 频道 Ai4 于 2025 年 8 月 24 日发布。视频内容为 LXILC 创始人 Alexandre Mogilamy 展示其开发的 Conclave AI Engine，介绍该多智能体推理平台在地缘政治兵棋推演、金融交易分析以及复杂问题辩论协作等领域的应用与潜力。 ## TL;DR Alexandre Mogilamy 详细演示了 Conclave AI Engine 如何打破传统单一模型与黑盒推理的局限，通过调动多达七个主流大语言模型开展两轮辩论与交叉评审，生成经典、替代及黑天鹅视角的综合解答。平台不仅在地缘政治危机与跨国制药企业竞争模拟中实现了分钟级的多角色战略兵棋推演，还能在金融交易中快速解析海量数据为散户赋能。最后，Alexandre 分享了自己作为全职调酒师利用业余时间仅耗时四个半月独立完成系统开发的经历，并展望了司法、医疗与多角色伴侣等未来应用场景。 ## Atmosphere & style 整个演讲节奏明快、充满激情与感染力。演讲者以极其自信且带有些许法式幽默的口吻侃侃而谈，巧妙地将调酒经验与严密的多参数 Prompt 工程相类比，生动化解了复杂技术的门槛感。结尾揭示自身非科班背景与业余独立开发的经历更是将现场情绪推向高潮，既务实硬核又极具草根极客式的鼓舞力量。 ## Key takeaways - [00:04](https://www.youtube.com/watch?v=p47a2t060a8&t=4s) 市面上许多 AI 系统内部多智能体协作流程封闭，Conclave 通过引入七个主流大模型并透视两轮协作过程，打破了传统黑盒模式。 - [00:04](https://www.youtube.com/watch?v=p47a2t060a8&t=4s) Conclave 采用双轮结构化推理：第一轮产出改良、替代、黑天鹅及颠覆性视角，第二轮由其他模型交叉评审并深化润色。 - [00:04](https://www.youtube.com/watch?v=p47a2t060a8&t=4s) 多模型协同辩论展现了极高效率，例如在应对复杂超级智能议题时，全流程多维度生成仅耗时 39 秒。 - [02:28](https://www.youtube.com/watch?v=p47a2t060a8&t=148s) Conclave 兵棋推演将地缘政治危机（如南海局势）解构成五角大楼、中国、美国国会等多方阵营，通过定制 Prompt 和交战规则进行博弈模拟。 - [02:28](https://www.youtube.com/watch?v=p47a2t060a8&t=148s) 兵棋推演架构可无缝迁移至商业竞争，支持对药企自身、三大竞对、监管审批及媒体舆论传导进行全方位多主体建模。 - [02:28](https://www.youtube.com/watch?v=p47a2t060a8&t=148s) 创始人以调酒师经验类比，指出高质量 AI 输出的核心在于精准输入多维参数，Conclave 在底层承担了连接海量现实变量与 Prompt 的桥梁角色。 - [02:28](https://www.youtube.com/watch?v=p47a2t060a8&t=148s) 企业级复杂推演在 4 分 30 秒内即可完成，其敏捷度与方案深度对标传统咨询巨头（如麦肯锡）的输出水准。 - [05:40](https://www.youtube.com/watch?v=p47a2t060a8&t=340s) Conclave Trading 支持吞吐高达 600 万 Token 的金融数据集，能在 1 至 2 分钟内为散户提供媲美顶级对冲基金的系统化量化与期权策略。 - [05:40](https://www.youtube.com/watch?v=p47a2t060a8&t=340s) 未来交易引擎计划引入财报会议管理层肢体语言视频分析与供应链卫星图像，进一步拓宽多模态输入维度。 - [05:40](https://www.youtube.com/watch?v=p47a2t060a8&t=340s) Conclave 核心推理引擎具备高度模块化特性，可横向赋能司法法庭对抗、多角色情感伴侣、医疗科学论证及博彩赔率等多元垂直场景。 - [05:40](https://www.youtube.com/watch?v=p47a2t060a8&t=340s) 创始人非科班出身且兼职在酒店工作，仅耗时 4 个半月独立开发出系统，以此证明多智能体与 AI 落地构建的普适性潜力。 ## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=p47a2t060a8&t=4s) Conclave 多模型问答引擎：双轮协作与多视角对抗 Alexandre Mogilamy 登台向观众展示了他构建的多智能体协作平台——Conclave AI Engine。他开场便引导现场观众扫描大屏幕上的二维码，直接用手机实时体验平台应用，并提到自己在展会 454 号展位也部署了全套实时演示。Alexandre 解释道，当前市面上的主流 AI 往往是“单一问答”模式（即一个问题对应一个模型的一个回答），即便是声称使用了五个不同 Agent 的系统，其内部推理流程对用户而言也是完全封闭的黑盒，用户根本不知道答案是如何协同生成的。Conclave 则彻底改变了这种范式，引入了七个主流大语言模型共同处理一个复杂问题。 ```mermaid flowchart TD accTitle: Conclave 多模型协作工作流 accDescription: 展示 Conclave 引擎在两轮推理中如何利用 7 个不同 AI 模型进行分工、对抗与融合。 subgraph Round1["第一轮：初步生成与多元视角"] Q["用户输入问题"] --> M1["模型 1 & 2<br/>（如 Cohere / Mistral）<br/>生成基线与替代方案"] Q --> M2["模型 3 & 4<br/>（如 GPT / Grok）<br/>生成黑天鹅与颠覆性视角"] end subgraph Round2["第二轮：交叉评审与深化改进"] M1 --> R2["模型 5 & 6<br/>（如 Claude / Perplexity）<br/>交叉评审、迭代并优化最佳方案"] end subgraph Output["最终呈现"] R2 --> F1["经典/渐进式答案 (Classic / Ameliorated)"] R2 --> F2["创新替代答案 (Alternative)"] M2 --> F3["黑天鹅与颠覆性预测 (Black Swan / Subversive)"] end ``` 在 Conclave 的问答机制中，系统通过两轮（Round）结构化推理来输出具有深度梯度的回答矩阵： - **第一轮（Round 1）**：四个模型负责提出初步的改良型与替代型答案；另外两个模型则跳出常规思维，分别生成“黑天鹅（Black Swan）”式的未来预测与“颠覆性/颠覆式（Disruptive / Subversive）”视角。 - **第二轮（Round 2）**：上一轮产出的改良与替代方案会进入第二轮，由其他模型（如 Claude、Perplexity 等）进行交叉评审、筛选并进一步润色改进。 以“当 AI 在所有领域超越人类智能时会发生什么？”这一问题为例，系统的“经典回答”是由 Cohere 在第一轮提出初稿、再由 Claude 在第二轮进行深度改进；“替代回答”则由 Mistral 起草、经由 Perplexity 优化生成；同时 GPT 贡献了黑天鹅预案，Grok 提供了极具冲击力的颠覆性答案。整个多模型辩论与生成全过程仅耗时 39 秒，用户不仅能看到最终结论，还能完全透视答案演变与模型交互的每一个环节。 --- ### [02:28](https://www.youtube.com/watch?v=p47a2t060a8&t=148s) Conclave 战略兵棋推演与企业竞争模拟 Alexandre 进而展示了该引擎在复杂战略兵棋推演（War Gaming）中的应用。在模拟“南海局势升级（South China Sea escalation）”的地缘政治危机中，系统将不同的 AI 实体分别分配给不同阵营的行动方：五角大楼（Pentagon）、中国、日本、美国国会以及华尔街。每一个角色都有专门定制的高精度 Prompt 体系，涵盖态势感知、战略目标、关键参数、任务清单、盟友协同机制以及交战规则（Rules of Engagement）。推演启动后，各 AI 角色根据各自利益与规则进行实时交互决策，系统最终输出包含各方决策要点的综合简报（包含 380 字符的核心摘要及 3,874 字符的深度推演明细）。 推演架构同样被 Alexandre 迁移到了商业竞争领域。全球顶尖的跨国制药巨头曾委托他为其定制战略推演： - **全方位博弈建模**：系统同时对该企业自身、前三大竞争对手、医药监管机构（如 FDA/EMA 审批动态）以及主流媒体进行多维度建模。 - **媒体舆论传导模拟**：推演特别考量了财经媒体的立场与影响力（例如彭博社 Bloomberg 与该药企及竞品的潜在利益关联、头条新闻的议题设置机制以及其他记者的跟风反应）。 谈及如何让 AI 胜任如此复杂的任务时，Alexandre 分享了自己作为调酒师（Bartender）的切身体会。朋友曾对他说“AI 根本调不出好鸡尾酒”，他反驳道：“关键在于你有没有提供精准的参数。”如果明确输入柠檬汁的酸度（如 0.7）、糖浆比例、冰块规格、玻璃器皿类型与特定手法，AI 就能给出极致精准的配方。Alexandre 直言自己在 Conclave 中扮演的就是连接复杂现实与 AI 的“中间人（Middleman）”——用户往往不知道如何把海量变量转化为高质量 Prompt，而 Conclave 引擎负责在底层建立严密的多参数交互架构。 - **推演生成效率**：该药企完整复杂场景的推演全流程仅需 4 分 30 秒。 - **咨询对标**：Alexandre 幽默自嘲道：“这套 V1 版本的概念验证（PoC）产出已经逼近麦肯锡（McKinsey）咨询公司的成果。这听起来可能有点狂妄，但抱歉，我是法国人。”他指出传统咨询巨头行动迟缓且策略僵化，而 AI 驱动的多智能体推演具备极高的敏捷度。 --- ### [05:40](https://www.youtube.com/watch?v=p47a2t060a8&t=340s) Conclave 交易分析引擎与多领域未来版图 在金融交易场景中，Conclave Trading 展现了强大的多模态与多智能体数据吞吐能力。系统能够直接接入端到端市场行情数据（单次处理高达 600 万 Token 的庞大金融数据集），由后台多个 AI 分工解析图表走势、技术指标与量化信号，并在 1 分 10 秒至 2 分钟内生成涵盖短期、中期与长期的多维度交易策略。对于更复杂的深度研判，分析过程也仅需约 3 分 30 秒。Alexandre 强调，系统旨在打破顶级对冲基金（Hedge Funds）与普通散户之间的信息壁垒，让散户在交易股票或期权时，不再依靠主观“直觉”盲目亏损，而是能够系统化评估期权行权价（Strike）、到期日（Expiration）等多重变量。 ```mermaid flowchart LR accTitle: Conclave 多领域生态拓展架构 accDescription: Conclave 核心引擎向交易、司法、医疗、伴侣和体育博彩等多个垂类场景的横向赋能。 Engine["Conclave 核心推理引擎<br/>(Modular Multi-Agent Core)"] Engine --> Trad["金融交易 (Trading)<br/>行情图表 / 6M Token 数据 / 卫星与肢体分析"] Engine --> WarG["战略兵棋 (War Gaming)<br/>地缘政治危机 / 商业竞对推演"] Engine --> Legal["司法模拟 (Justice & Legal)<br/>法官 / 辩护律师 / 检方博弈"] Engine --> Comp["多角色伴侣 (AI Companions)<br/>5 种人格同时提供多维反馈"] Engine --> MedSci["医疗与科学 (Medical & Science)<br/>医药策略 / 科研论证"] Engine --> Bet["概率分析 (Betting & Odds)<br/>胜率计算 / 综合赔率解析"] ``` 展望未来，Alexandre 规划了 Conclave 引擎的多个垂直延伸方向： - **交易进阶维度**：引入管理层财报会议的肢体语言（Body Language）视频分析与工厂供应链的卫星图像（Satellite Imagery）监测。 - **多模型情感伴侣（Multi-AI Companions）**：用户发送一条信息，由五个不同性格特征的模型同时回应（如密友、知己、导师、唱反调者等），全方位满足情感与决策需求。 - **司法与法律博弈模拟（Justice & Legal）**：模拟法庭环境，提前推演己方主张提出后，辩方律师会如何反驳、法官将如何裁决。 - **医疗（Medical）、前沿科学（Science）与体育博彩赔率分析（Betting & Odds）**。 在演讲尾声，Alexandre 透露了极具感染力的个人背景：他并非技术科班出身，目前在拉斯维加斯百乐宫酒店（Bellagio）做着全职工作，利用业余时间从零开始独立开发，仅耗时 4 个半月就完成了整个 Conclave 系统的构建。他用这段经历鼓励全场：“每个人都能做到这一点。”他再次感谢观众的时间，并邀请大家前往 454 号展位使用现场平板设备亲自上手体验各项功能。 ## Notable quotes > [00:04](https://www.youtube.com/watch?v=p47a2t060a8&t=4s) 一个问题对应一个模型的一个回答，即便声称使用了五个不同 Agent 的系统，其内部推理流程对用户而言也是完全封闭的黑盒，用户根本不知道答案是如何协同生成的。 — Alexandre Mogilamy > [00:04](https://www.youtube.com/watch?v=p47a2t060a8&t=4s) 当 AI 在所有领域超越人类智能时会发生什么？ — Alexandre Mogilamy > [02:28](https://www.youtube.com/watch?v=p47a2t060a8&t=148s) 关键在于你有没有提供精准的参数。 — Alexandre Mogilamy > [02:28](https://www.youtube.com/watch?v=p47a2t060a8&t=148s) 用户往往不知道如何把海量变量转化为高质量 Prompt，而 Conclave 引擎负责在底层建立严密的多参数交互架构。 — Alexandre Mogilamy > [02:28](https://www.youtube.com/watch?v=p47a2t060a8&t=148s) 这套 V1 版本的概念验证（PoC）产出已经逼近麦肯锡（McKinsey）咨询公司的成果。这听起来可能有点狂妄，但抱歉，我是法国人。 — Alexandre Mogilamy > [05:40](https://www.youtube.com/watch?v=p47a2t060a8&t=340s) 系统旨在打破顶级对冲基金（Hedge Funds）与普通散户之间的信息壁垒，让散户在交易股票或期权时，不再依靠主观“直觉”盲目亏损，而是能够系统化评估期权行权价（Strike）、到期日（Expiration）等多重变量。 — Alexandre Mogilamy > [05:40](https://www.youtube.com/watch?v=p47a2t060a8&t=340s) 用户发送一条信息，由五个不同性格特征的模型同时回应，全方位满足情感与决策需求。 — Alexandre Mogilamy > [05:40](https://www.youtube.com/watch?v=p47a2t060a8&t=340s) 每个人都能做到这一点。 — Alexandre Mogilamy ## Mentioned Alexandre Mogilamy, Conclave AI Engine, LXILC, Legacy eXperience Initiative Limited Company, Ai4, Cohere, Mistral, GPT, Grok, Claude, Perplexity, 五角大楼 (Pentagon), 美国国会 (US Congress), 华尔街 (Wall Street), FDA, EMA, 彭博社 (Bloomberg), 麦肯锡 (McKinsey), 百乐宫酒店 (Bellagio) ## Video description Alexandre Mogilamy, FOUNDER, Legacy eXperience Initiative Limited Company (LXILC) Abstracts What if you could simulate and manage the most critical geopolitical crises, test your strategic decisions, and anticipate the unpredictable—in real-time, powered by AI? 1. Conclave War Gaming Imagine facing a sudden escalation in the South China Sea or managing a diplomatic standoff between global powers. Conclave War Gaming places multiple specialized AI agents into realistic geopolitical and military scenarios, each role-playing distinct actors (Pentagon, Congress, Foreign Governments, Market Analysts). Engage in immersive simulations, uncover hidden risks, and test strategic options under pressure. Conclave War Gaming isn’t a simulation—it's strategic foresight powered by advanced, collaborative AI. 2. Conclave Trading Markets never sleep—neither does Conclave Trading. Our advanced AI engine analyzes stocks, identifies signals, anticipates trends, and delivers sophisticated financial insights usually accessible only to top-tier hedge funds. It transforms massive financial datasets into actionable strategies by leveraging multi-agent reasoning, combining rigorous technical analysis with intelligent predictive modeling. With Conclave Trading, navigate uncertainty, anticipate market moves, and seize opportunities ahead of the curve. 3. The Conclave (Questions) Complex questions deserve more than a single answer. Conclave leverages multiple advanced AI agents—including GPT, Claude, Perplexity, Grok, and more—to debate, reason, challenge each other, and converge towards nuanced, insightful responses. Ask philosophical dilemmas, strategic business problems, or real-world questions, and watch the Conclave engine transform isolated answers into dynamic dialogues and thoughtful, collective insights. Powered by the Conclave AI Engine At the heart of all these applications is the Conclave Engine: a modular, multi-agent reasoning platform designed for complexity and uncertainty. It enables curated scenarios, intricate role assignments, and advanced logic interactions—delivering insights far beyond single-agent solutions. Conclave isn’t just AI answering questions—it's intelligence debating, strategizing, and evolving. Join us and experience the power of collaborative AI thinking. [Source](https://www.youtube.com/watch?v=p47a2t060a8)

Source quotes

[0:04] Alexandre Mogilamy: 一个问题对应一个模型的一个回答，即便声称使用了五个不同 Agent 的系统，其内部推理流程对用户而言也是完全封闭的黑盒，用户根本不知道答案是如何协同生成的。 [0:04] Alexandre Mogilamy: 当 AI 在所有领域超越人类智能时会发生什么？ [2:28] Alexandre Mogilamy: 关键在于你有没有提供精准的参数。 [2:28] Alexandre Mogilamy: 用户往往不知道如何把海量变量转化为高质量 Prompt，而 Conclave 引擎负责在底层建立严密的多参数交互架构。 [2:28] Alexandre Mogilamy: 这套 V1 版本的概念验证（PoC）产出已经逼近麦肯锡（McKinsey）咨询公司的成果。这听起来可能有点狂妄，但抱歉，我是法国人。 [5:40] Alexandre Mogilamy: 系统旨在打破顶级对冲基金（Hedge Funds）与普通散户之间的信息壁垒，让散户在交易股票或期权时，不再依靠主观“直觉”盲目亏损，而是能够系统化评估期权行权价（Strike）、到期日（Expiration）等多重变量。 [5:40] Alexandre Mogilamy: 用户发送一条信息，由五个不同性格特征的模型同时回应，全方位满足情感与决策需求。 [5:40] Alexandre Mogilamy: 每个人都能做到这一点。

Source quotes

[0:04] Alexandre Mogilamy: nobody know where the answer go or is create [2:23] Alexandre Mogilamy: AI today they can do everything [2:23] Alexandre Mogilamy: we just have to know to talk to them to ask the perfect question with all the data. What if you could simulate and manage the most critical geopolitical crises, test your strategic decisions, and anticipate the unpredictable—in real-time, powered by AI? Conclave War Gaming isn’t a simulation—it's strategic foresight powered by advanced, collaborative AI. Markets never sleep—neither does Conclave Trading. Complex questions deserve more than a single answer. Conclave isn’t just AI answering questions—it's intelligence debating, strategizing, and evolving.

Original transcript

[0:04] So I create an engine. The name is trunk [0:06] lab. But right now if you want you can [0:08] follow everything you're going to say [0:10] with the QR code and you have access to [0:12] your phone to the whole experience. [0:13] Everything is live. I have a stand the [0:17] 454 and everything is live as well. Oh [0:20] let me explain you what I'm doing and [0:23] pass. Okay. So the klav is an engine [0:27] using different AI. pretty simple. So [0:31] what people do currently so it's a [0:34] single AI one question one AI one answer [0:36] this is what you have with GPT and [0:38] everything recently different model as [0:42] Gemini grow they use different agents so [0:46] they say oh we use five different agents [0:48] the question go to five different agents [0:50] but nobody know where the answer go or [0:53] is create and they provide you final [0:54] answer what we doing with lab question [0:58] the clab you have one question seven AI [1:01] treat your question from the interface [1:04] you see the answer coming. So for AI [1:07] they give you some alternative answer [1:09] and amorate answer in the first round [1:11] and the two other they give you a black [1:13] swan futuristic answer and as well a [1:16] disruptive answer. We have a second [1:18] round when the four answer so [1:22] alternative and amidoritates go on the [1:24] next run and they are analyzed again by [1:26] the four AI. They selected the best AI [1:30] the best answer sorry and they improve [1:32] it and at the end you have all the [1:35] answer you can see how it was create so [1:38] this is an example what happened when AI [1:40] surpass human intelligence in all domain [1:43] so the first answer Gemini take this [1:45] answer you can see it was proposed by [1:48] coair in a run one and improved by load [1:50] in a run two this is the classic answer [1:53] after we have the surn answer is the [1:55] alternative answer so mistral in the [1:57] First round make it perplexity in the [1:59] second round improve it. [2:02] So at the end we give you the two more [2:04] answer. So GPT the black swan answer and [2:07] grow the submersive answer and you can [2:09] and it was generates in 39 seconds is [2:13] longer than a normal EI but you have a [2:15] multi talking together and at the end [2:18] you can have access to all the answer [2:20] they've been generates. So this is one [2:23] of the way I use my engine. This engine [2:26] sorry I can use in different way for [2:28] example the wargaming. So the war gaming [2:31] is the same thing. We use the same [2:33] engine different AI engine sorry and we [2:36] put them together. So we have a scenario [2:38] here the source China sees estellation [2:43] sorry for the word. So you need a [2:45] context as well. You need an event then [2:47] you provide to the EI. This is a shorter [2:50] version for you. So the actor, we have [2:52] the Pentagon, the China, the Japan, the [2:54] Congress and Wall Street. So they all [2:56] have some different perspective, some [2:59] different thing to do. Each AI they play [3:01] a different actor. Everything in AI is [3:04] the prompt. So I put you a shorter [3:07] prompt than I will put I can't put [3:10] everything. It's too long, but give you [3:11] an idea. So in your prompt you put the [3:13] situation the strategic objective the [3:16] key parameters the uh the mission the [3:19] alias coordination the rules of [3:21] engagements and it's just a short [3:22] version of the prompt the prompt is way [3:24] bigger usually but is an example what [3:27] you can do so each AI have a prompt they [3:30] know the scenario they can see or [3:32] interact each other and at the end this [3:35] is a resume than what happened in this [3:37] situation so we just have the the [3:39] headline the decision of the pentagon [3:41] the China, the Japan, the Congress, the [3:43] Wall Street, and you have 380 more [3:46] character summary and 3,874 additional [3:49] character at the end. So you can see [3:52] more in detail in the apps, but I have [3:54] to compress for you for you understand [3:56] what we're doing. So that's can be add [3:59] for different strategy. I've been [4:01] reached by a big company for [4:03] Pharmaceutic, one of the biggest in the [4:05] world. They won't do test strategy. So I [4:08] do the same. I take them company I check [4:10] the concurrenance one two three we check [4:13] the board like K they handle the [4:16] medicine we handle everything we handle [4:19] the media so which media so for example [4:21] for media which media is the most uh [4:24] able to talk about them who is the main [4:26] one is Bloomberg or Bloomberg interact [4:29] usually with this company if Bloomberg [4:31] have interest with other company when [4:33] Bloomberg talk if this headline about [4:35] this topic is not a headline what do the [4:38] other journalist So you have to take in [4:40] consideration this plus you ask to all [4:42] the AI you can what to take in [4:44] consideration you make your prompt is a [4:47] bit longer but AI today they can do [4:50] everything we just have to know to talk [4:52] to them to ask the perfect question with [4:55] all the datal little example I'm [4:57] bartender my friends say AI can't make a [5:00] cocktail I said to them did you ask them [5:03] they can use all the juice in the world [5:04] they can make their own juice they can [5:06] use their own fruits they can use [5:07] different glassware different ice they [5:10] can use a different method yeah but the [5:12] taste is not good okay did you told them [5:14] the acidity of your lemon juice 0.7 I [5:17] don't know or the simple zero did you [5:19] told them we oh no so make this [5:22] parameter make a question you will see [5:23] the answer so my job with lab I'm the [5:26] middleman you are here the AI is here [5:29] you want the best answer but you don't [5:31] know to ask a question it's too [5:32] complicated it's too many parameters we [5:34] create an engine I create an engine with [5:37] this AI and we can make it happen. So [5:40] the next thing the cont threading is [5:42] another example or I can use my my [5:45] technology. So it's pretty simple. So [5:48] this is what happened at the end. So we [5:50] have an AI of trading is EOD. He provide [5:53] us six million of token for example. So [5:56] I have a different AI behind they [5:58] combine that and even I don't tell six [6:00] million of token is too large. This CI [6:02] they take this information different AI [6:04] they read they check the charts the [6:07] final uh everything you have in [6:10] techation for trading uh and they [6:13] provide you a final answer of long-term [6:15] short-term midterm the version available [6:17] right now my is free right now I will [6:20] have to charge the people it's pretty [6:22] simple it didn't fight with edge front [6:24] or anything but it can so if we add [6:28] after the genetic trading we can Add for [6:31] example the body language analyze we can [6:34] add a manifactor satellite picture I do [6:37] it by myself in a different way it can [6:39] be add and in 1 minute 30 2 minutes you [6:43] can have this result it was generated in [6:46] 1 minute and 10 seconds it's a short [6:49] it's a shorter one but you can go bigger [6:51] and faster in the ah fonts level the [6:54] same I forget to tell you so everything [6:56] you can try on the apps you have the [6:57] link or at the end I give you back the [6:58] link um for what we can do previously [7:02] for the strategy. It can be for some [7:04] some company at the end when I put the [7:06] full scenario for this company the time [7:08] is 4 minutes and a half to generate the [7:11] full scenario it's it's my V1 a P for [7:14] them a proof of concept it's already [7:16] close to MC consulting product why [7:21] because this kind of company they don't [7:22] move they keep the same strategy they [7:24] don't use the eye they're not flexible I [7:26] am you look pretentious but I'm French [7:28] sorry for that so now we're going to see [7:31] so the trading model we come back to [7:34] trading. It's what he give to you. So [7:35] the idea it's any people they won't do [7:38] trading they don't have the tools to [7:39] trade properly. So with these tools in [7:42] two minutes for the short version maybe [7:44] 3 minutes and a half for the very detail [7:46] version will be available. You have all [7:48] the information about trading close to [7:49] the edge funds instead of look chart and [7:52] have a feeling and losing your money in [7:54] two trade. You can technically revalize [7:56] for the hedge funds and for the net [7:59] version if you do option or stock you [8:01] can put in it. If you have option you [8:03] have expiration you have strike uh you [8:06] have many parameter then we can put in [8:08] the system even you already if you buy [8:10] already or not yet. So all of this the [8:12] goal is to give to everyone the best [8:15] tool to success in a different field. I [8:18] hope it's clear. So basically klav today [8:21] this is the three module I show to you [8:23] but it can be extend like why not you [8:26] find a bet you find a fight we analyze [8:29] all the odds and we find the way to find [8:31] the answer companion so everybody talk [8:34] to AI right now but imagine you talk to [8:36] five AI one message one is the friend [8:39] one is the confidence one give you [8:41] advice once is the bad one they say oh [8:43] you do bad it can be improve as well [8:45] with the same system the system can do [8:49] cryptocurrency insteading [8:52] uh the justice your lawyers you can set [8:55] up if you make a decision if you pro if [8:58] you come with an argument what the [9:00] defender will do what the judge going to [9:02] say and you can put the parameters of [9:04] all of them uh the klab medical it's a [9:07] very big business and it can help [9:09] everyone and the klab science so [9:11] basically you can come with a need or in [9:14] my engine and my I my knowledge I help [9:16] you to build something or you use what I [9:19] build already. So basically this is what [9:22] I create what I do. I do that in like [9:24] four months and a half. I have a [9:26] full-time job at the Bellagio. I'm not [9:28] from the tech and I do that from [9:30] scratch. So everybody can do it. Take a [9:32] lot of time. I'm try but I'm going. So [9:36] this is a QR code. If you want try again [9:38] the experience you can have access to [9:39] the three I told you so far. It's free [9:42] and you have my contact on this page. [9:44] I'm at the stand 545 454 sorry I going [9:48] to go back right now because I work by [9:49] myself in everything tomorrow I will be [9:51] there and I have some tablet you can try [9:53] as well the experience so I hope you [9:56] like what I say I hope uh I convince you [9:59] maybe and thank you again for your time [10:01] it's really appreciate because it's a [10:03] lot of thing to do today and I hope see [10:05] you to my stand or reach me in any way [10:07] you want thank you very