AI Market Intelligence is an advanced, multi-agent financial analytics platform powered by Google Gemini 2.0.
It enables researchers, traders, asset managers, and risk professionals to:
- Generate benchmark-aware investment insights
- Analyze global equities
- Compute quantitative risk metrics
- Explore markets using interactive AI explanations
The platform integrates real market data, multi-agent reasoning, quantitative finance, and interactive visual analytics into a single institutional-grade workflow.
- Overview
- Screenshots
- Key Features
- Multi-Agent Architecture
- System Capabilities
- Quantitative Analytics
- Tab-by-Tab Breakdown
- Data Sources
- Installation & Setup
- Tech Stack
- Target Users
- Future Enhancements
AI Market Intelligence is a decision-support platform combining:
- 🧩 Multi-agent AI reasoning
- 📊 Benchmark-aware financial analysis
- 📈 Quantitative market analytics
- 💬 Real-time sentiment extraction
- 🌎 Global market data
- 💼 Portfolio optimization
- 🤝 Interactive AI insights
It is designed to help professionals make faster, smarter, and data-driven decisions in financial markets.
Six highly specialized AI agents independently analyze different domains:
| Agent | Role |
|---|---|
| 🧠 MarketAnalystAgent | Market trends, volatility, momentum & regime detection |
| 🏢 CompanyResearchAgent | Fundamentals, earnings, catalysts, valuation |
| 💬 SentimentAgent | Region-aware news sentiment from live RSS feeds |
| ⚖️ RiskAnalystAgent | VaR, CVaR, beta, drawdown, correlations, stress tests |
| 📊 PortfolioStrategistAgent | Allocation strategies: equal weight / risk parity / momentum |
| 👨💼 TeamLeadAgent | Synthesizes all insights into a final benchmark-aware report |
- 🌎 Multi-country ticker normalization
- 📈 Regional benchmark selection
- 🏛️ Exchange suffix auto-application
- 📰 Region-aware news in multiple languages
Includes institutional-grade risk & performance analytics:
- Rolling Volatility
- Rolling Beta (vs benchmark)
- Correlation Heatmap
- Maximum Drawdown
- Alpha & Tracking Error
- Value-at-Risk (VaR)
- Conditional VaR (CVaR / Expected Shortfall)
- Sortino Ratio
- Sector Composition Approximation
Stress Testing:
- ⚡ Shock-based stress test
- 📉 Historical analog crash scenarios
All insights, charts, and metrics are contextualized against benchmark indices, such as:
S&P 500, NASDAQ 100, Dow Jones, FTSE 100, Nikkei 225, Nifty 50, Euro Stoxx 50
Users can ask Gemini:
"Explain this chart to me"
And receive:
- Observations
- Risk insights
- Anomalies
- Trend interpretation
- Recommendations
Supports multiple allocation frameworks:
- Equal-Weight
- Risk-Parity (inverse volatility)
- Momentum Tilt
- Custom Inputs
Includes expected risk/return interpretation for each strategy.
A conversational interface to ask about:
- Market questions
- Company comparisons
- Risk explanations
- Benchmark analysis
- Financial definitions
- Portfolio construction
The system follows a multi-agent orchestration pipeline where each specialized AI agent contributes domain-specific intelligence to the final report.
Each Agent Receives
- 🧹 Cleaned and pre-processed data
- 📈 Historical price series
- 📊 Computed returns and volatility metrics
- 🌍 Regional sentiment information
- 🧮 Benchmark-adjusted metrics
The TeamLeadAgent consolidates all agent outputs, applies benchmark awareness, and generates a final institutional-grade report that includes:
- 🧠 Synthesized insights
- 📉 Quantitative performance metrics
- ⚖️ Risk evaluation
- 💼 Portfolio recommendations
- 💬 Interactive explanations via Gemini
- ✔ Global equity analysis
- ✔ Benchmark-relative performance
- ✔ Rolling factor analytics
- ✔ Quantitative risk modeling
- ✔ AI-assisted chart interpretation
- ✔ Portfolio design & optimization
- ✔ News sentiment classification
- ✔ Multi-agent chain-of-thought synthesis
- ✔ Rich visualization suite (powered by Matplotlib & Seaborn)
| Metric | Description |
|---|---|
| VaR (Value-at-Risk) | Worst expected loss at a given confidence level |
| CVaR (Expected Shortfall) | Average loss in tail-risk events |
| Rolling Beta | Benchmark sensitivity over time |
| Alpha | Outperformance vs benchmark |
| Tracking Error | Deviation from benchmark |
| Correlation Matrix | Inter-asset co-movements |
| Drawdown Curve | Historical peak-to-trough dynamics |
| Stress Test | Shock simulation & historical crash modeling |
| Sector Exposure Approximation | Equal-weighted sector inference |
Beginner-friendly introduction & glossary.
- 📈 Price charts
- 📊 Cumulative returns
- 🧠 MarketAnalystAgent insights
- 🏢 Company metadata
- 💰 Financials
- 📰 News sentiment
- 🧾 CompanyResearchAgent results
- 📉 Volatility
- 🔥 Correlation heatmaps
- ⚖️ Rolling beta
- 🧮 CVaR & VaR
- 💥 Stress tests
- 👁️ RiskAnalystAgent insights
- 💬 Interactive Gemini explanations
- 📊 Multi-chart insights
- ⚖️ Equal-weight
- 📈 Risk parity
- 🚀 Momentum tilt
- 🤖 PortfolioStrategistAgent recommendations
- 🗣️ Context-aware question routing
- 🤝 Multi-agent responses
- 🧾 TeamLeadAgent multi-agent report
- 📘 Benchmark-aware synthesis
| Source | Usage |
|---|---|
| Yahoo Finance | Prices, metadata, and returns |
| RSS News Feeds | Region-aware news sentiment |
| Google Gemini 2.0 | AI reasoning, synthesis, and explanations |
Clone the repository and install the required dependencies:
pip install -r requirements.txtEnsure you add your Google API key as an environment variable
GOOGLE_API_KEY="your_api_key"
Run the Streamlit application
streamlit run app.py
| Component | Technology |
|---|---|
| 🖥️ Frontend | Streamlit |
| ⚙️ Backend | Python |
| 🧠 AI Models | Google Gemini 2.0 Flash |
| 💾 Data Sources | Yahoo Finance + RSS Feeds |
| 📊 Visualization | Matplotlib / Seaborn |
| 🤖 Orchestration | Custom Multi-Agent Framework |
The platform is designed for:
- 🧾 Equity Researchers
- 💹 Traders
- 💼 Asset Managers
- ⚖️ Risk Managers
- 📈 Quantitative Analysts
- 🎓 Finance Students
- 💻 FinTech Developers
- 📉 Monte Carlo return simulation
- 🧮 Efficient frontier (mean-variance optimization)
- 📚 Factor model integration (Fama–French)
- 💰 ETF + Crypto asset class support
- 💾 Persistent user sessions