Stock Market Sentiment Analyzer

An automated AI-powered system that fetches financial news daily, analyzes sentiment using Claude AI, and delivers personalized investment insights via a morning email digest and a real-time dashboard.

Built by Arya Patel.

---

What It Does

• Fetches financial news every morning at 7 AM via NewsAPI
• Tracks your personal watchlist + top 10 most actively traded stocks
• Uses Claude AI to score each article as bullish, bearish, or neutral
• Stores results in Supabase for historical trend analysis
• Sends a morning email digest with investment signals before the market opens
• Displays a real-time React dashboard showing 30-day sentiment trends

---

Tech Stack

Layer	Technology
Backend	Python 3.10, FastAPI
Scheduler	Apache Airflow
AI	Anthropic Claude API
News	NewsAPI + Financial Modeling Prep
Database	Supabase (PostgreSQL)
Frontend	React, Chart.js, Tailwind CSS
Email	Gmail SMTP (dev), SendGrid (prod)
Containerization	Docker
CI/CD	GitHub Actions

---

Project Structure

stock-market/
├── dags/                   # Airflow DAG (daily pipeline)
├── src/
│   └── sentiment_analyzer/
│       ├── news/           # fetch_financial_news, fetch_top_tickers
│       ├── sentiment/      # analyze_sentiment (Claude API)
│       ├── database/       # store_sentiment_result (Supabase)
│       ├── email/          # send_email_digest
│       ├── api/            # FastAPI endpoints
│       └── utils/          # logger, config
├── tests/                  # Unit tests (pytest)
├── .github/workflows/      # CI/CD (GitHub Actions)
├── Dockerfile
├── docker-compose.yaml
└── requirements.txt

---

Setup

1. Clone the repo

git clone https://github.com/apate476/Stock-Market.git
cd Stock-Market

2. Create a virtual environment

python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate

3. Install dependencies

pip install -r requirements.txt

4. Configure environment variables

cp .env.example .env

Fill in your keys in .env:

ANTHROPIC_API_KEY=
NEWS_API_KEY=
FMP_API_KEY=
SUPABASE_URL=
SUPABASE_SERVICE_KEY=
EMAIL_SENDER=
EMAIL_PASSWORD=

5. Run tests

pytest tests/ --cov=src --cov-report=term-missing

6. Start the API

uvicorn src.sentiment_analyzer.api.main:app --reload

---

Pipeline Flow

fetch_financial_news → analyze_sentiment → store_sentiment_result → send_email_digest

Each step is an Airflow task. Data passes between tasks via XCom.

---

Personal Watchlist

Currently tracking: NVDA, SPCX, MSFT, BRK-B, JPM + top 10 most active stocks fetched daily from Financial Modeling Prep.

---

License

MIT License — see LICENSE for details.