12 AI Agents. 1 Ticker. Real-Time Debate. Consensus Decision.
--- ## What is this? **AI Quant Fund** is a multi-agent trading analysis system where **12 specialized AI agents** debate and analyze stocks in real-time โ just like a Wall Street trading floor, but powered entirely by AI. ``` You enter a ticker โ 12 AI agents activate โ They debate โ You get a consensus ``` ### The 12 Agents | Phase | Agent | Role | |-------|-------|------| | ๐ Intelligence | Market Analyst | Technical analysis, price action, volume patterns | | ๐ Intelligence | Social Analyst | Social media sentiment, retail trader mood | | ๐ Intelligence | News Analyst | Breaking news, SEC filings, macro events | | ๐ Intelligence | Fundamentals Analyst | Revenue, P/E, balance sheet, cash flow | | ๐ฅ Debate | Bull Researcher | Makes the bullish case with evidence | | ๐ฅ Debate | Bear Researcher | Makes the bearish case with evidence | | ๐ฅ Debate | Aggressive Debater | Pushes for high-conviction trades | | ๐ฅ Debate | Conservative Debater | Argues for caution and risk management | | ๐ฅ Debate | Neutral Debater | Weighs both sides objectively | | โก Decision | Trader | Proposes the trade (buy/hold/sell, size, timing) | | ๐ก๏ธ Risk | Risk Manager | **VETO POWER** โ can block any trade | | ๐ฏ Final | Investment Judge | Final verdict after all debates | ### How It Works ``` โโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโ โ INTELLIGENCE โ โ โ DEBATE โ โ โ TRADING โ โ โ DECISION โ โ โ โ โ โ โ โ โ โ 4 Analysts โ โ Bull vs Bear โ โ Trader โ โ Risk Manager โ โ gather data โ โ 3 Debaters โ โ proposes โ โ Judge rules โ โโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโ ``` --- ## Quick Start ### Install ```bash pip install ai-quant-agents ``` ### Usage โ 3 Lines of Code ```python from ai_quant_agents import QuantClient client = QuantClient() # connects to wss://dream.hmyk.ai/ws/live # Analyze any US stock or China A-share result = client.analyze("NVDA") print(result.decision) # "BUY" / "HOLD" / "SELL" print(result.confidence) # 0.85 print(result.consensus) # {"buy": 8, "hold": 3, "sell": 1} ``` ### Live Streaming โ Watch Agents Debate ```python from ai_quant_agents import QuantClient client = QuantClient() # Stream real-time agent messages for msg in client.stream("TSLA"): print(f"[{msg.speaker}] {msg.message}") # [Market Analyst] TSLA showing bullish divergence on RSI... # [Bear Researcher] However, delivery numbers missed by 12%... # [Risk Manager] Position size should not exceed 3% of portfolio... ``` ### China A-Share Support ๐จ๐ณ ```python result = client.analyze("600519") # ่ดตๅท่ ๅฐ result = client.analyze("000001") # ๅนณๅฎ้ถ่ก result = client.analyze("ๆฏไบ่ฟช") # Chinese name supported ``` --- ## Example Output ```json { "ticker": "NVDA", "decision": "BUY", "confidence": 0.85, "consensus": {"buy": 8, "hold": 3, "sell": 1}, "risk_approved": true, "key_reasons": [ "Strong data center revenue growth +120% YoY", "Blackwell GPU demand exceeding supply", "Institutional accumulation detected" ], "risk_warnings": [ "Valuation stretched at 45x forward P/E", "China export restrictions uncertainty" ], "suggested_action": { "entry": "$142-145", "stop_loss": "$132", "target": "$165", "position_size": "2-3% of portfolio" } } ``` --- ## ๐ด Try the Full Live Experience The demo SDK gives you the **analysis result**. But the real magic is watching **12 agents debate in real-time** in our cyberpunk trading room: **What you get in the full version:** - ๐ด Real-time streaming debate (watch agents think out loud) - ๐ Live market data ticker (US + China A-shares) - ๐ฏ Historical analysis archive - ๐ฑ Mobile responsive UI - ๐ English + Chinese bilingual --- ## Architecture Built on [TradingAgents](https://github.com/TauricResearch/TradingAgents) (Apache 2.0) by Tauric Research. Extended with: - China A-share market support (Tushare data) - Real-time WebSocket live streaming - Hybrid LLM (GPT + Gemini) for speed + depth - Cyberpunk UI trading room --- ## Disclaimer โ ๏ธ **NFA โ Not Financial Advice.** This tool is for research and entertainment purposes only. AI analysis should not be used as the sole basis for investment decisions. Always do your own research. Past performance does not guarantee future results. --- ## License Apache License 2.0 โ See [LICENSE](LICENSE) for details. Built with โค๏ธ by [DEMAND AI](https://dream.hmyk.ai)