# Rm Judicaël

- **Event:** [Meta Global AI Developer Hackathon](https://cerebralvalley.ai/e/meta-ai-developer-hackathon)
- **When:** Oct 26, 12:00 PM – Nov 4, 9:00 PM (EST)
- **Where:** Online
- **Team:** [Geoffroy Judicael RAMIELSON](https://cerebralvalley.ai/u/Judicael)
- **GitHub:** https://github.com/rmjudicael/jyntra-ai---financial-market-intelligence-platform
- **Demo video:** https://youtu.be/wIKQPydzHrI
- **Gallery:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/257

JYNTRA AI is a multimodal financial intelligence platform designed to help traders analyze financial markets, understand charts, and assess risks before making decisions. It addresses a real challenge: market information is complex and scattered, while trading signals often lack context and clear explanations. JYNTRA AI brings these elements together in one environment, prioritizing structured analysis over blind signals.

The platform uses real market data from sources including Binance Spot, Frankfurter (ECB exchange rates), and Yahoo Finance. It supports chart analysis from MT4, MT5, and TradingView, using market structure and Smart Money Concepts (SMC), including Order Blocks, Fair Value Gaps, liquidity, support and resistance, BOS, and CHoCH. Its analytical approach separates three levels: observable facts, their interpretation, and potential market scenarios.

At its core, JYNTRA AI features six specialized AI agents: Technical Analyst, Market Research Analyst, Risk Analyst, Strategy Analyst, Red-Team Analyst, and Lead Analyst. They examine market conditions from different perspectives, challenge assumptions, and synthesize their findings into a BUY, SELL, or NO TRADE assessment. Each decision can include an entry zone, stop-loss, take-profit, and risk-to-reward ratio, making the analysis more transparent and risk-aware.

Meta technologies power this architecture. Muse Spark and Muse Wonder support reasoning, analysis, and visual interpretation. SAM 3.1 contributes to detecting and segmenting relevant chart regions. Muse Image generates financial visuals, while Muse Voice Transcribe enables voice-processing capabilities. A centralized server-side engine orchestrates model selection, provider routing, and execution monitoring, tracking models, latency, and processing status. It also includes fallback mechanisms and a Meta-only mode that prevents responses from alternative providers from being misrepresented as Meta-generated.

Beyond market analysis, JYNTRA AI includes JYNTRACK, a probabilistic research environment for testing strategies and tracking results; Battle / War Room, for comparing visual proposals with structured analyses; LYNARA AI, an intelligent assistant; and Workflow Studio and an integrated Academy.

JYNTRA AI does not promise profits or replace human judgment. Its purpose is to make financial analysis more accessible, systematic, and transparent by transforming complex market data and visual information into structured, actionable insights.

JYNTRA AI - Your AI. Your Edge.
Individually designed and developed in Madagascar, with a vision for global technological impact.

## More from Meta Global AI Developer Hackathon

- [rm-rf](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/254)
- [DraftMark](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/255)
- [Mrinal Desai](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/256)
- [Dcoder](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/258)
- [Kevin](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/259)
- [Disclosed · 披露了](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/260)

---

Markdown version of https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/257. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt