FinAIc - Splintered information. Sharp decisions.

AI-powered stock insight toolkit. Assembling a mosaic of market signals.

---

📋 Table of Contents

• Overview
• Features
• Tech Stack
• Installation
• Usage
• Example Output
• Ideal Use Cases
• License
• Contributing

---

Overview

FinAIc is a side project born out of curiosity for markets, data, and machine learning. It blends AI-powered sentiment analysis, basic fundamental scoring, and analyst rating trends to spit out rough buy/hold/sell signals. It pulls in real-time news, key financial ratios, and aggregated Wall Street opinions to surface patterns and spark ideas. Whether you're a trader, a quant, or just someone who enjoys tinkering with financial data, this toolkit is meant to be explored, not blindly followed.

⚠️ Disclaimer: This is not financial advice. It’s a hobby project. If the code says “Buy,” don’t take it personally — or literally.

The codebase has been cleaned up and documented with AI assistance to make it easier to read, tweak, and build on. Fork it, break it, improve it — all welcome.

---

Features

• Real-Time News Sentiment Analysis
  Fetches headlines and summaries from Finnhub, then uses an Ollama NLP model to assign a numeric sentiment score.

• Fundamental Data Scoring
  Calculates health metrics such as current ratio, sales growth, net margin, and beta to gauge financial strength.

• Analyst Recommendation Trends
  Aggregates buy/hold/sell ratings from multiple analysts for a wider perspective on market consensus.

• Multi-Signal Recommendation Engine
  Weights sentiment, fundamentals, and analyst scores to generate a unified recommendation.

• Exportable Results
  Outputs detailed .txt reports and summary .xlsx files for deeper research or record-keeping.

---

Tech Stack

• Language: Python 3.x

• Core Libraries:

  • pandas — data manipulation and Excel export
  • finnhub-python — market data API client
  • ollama — AI/NLP sentiment scoring
  • re — text cleaning & preprocessing

• APIs:

  • Finnhub.io for news and fundamentals (but you can modify it for every other API)
  • Ollama for local or server-based LLM-driven sentiment

---

Installation

# Clone the repository
git clone https://github.com/<your-username>/FinAIc.git
cd FinAIc

# Install dependencies
pip install -r requirements.txt

---

Usage

1. Obtain a free API key from Finnhub.io.
2. Set your Finnhub API key in the script.
3. Ensure Ollama is installed and your chosen model (e.g., gemma3:27b) is running.
4. Run the analysis.
5. Check your output files in the project folder:
   • analysis_result_<symbol>_<enddate>.txt
   • full_analysis_<enddate>.xlsx

The example files follow the naming convention shown above and can be found in the main repository folder.

---

Example Output

🧮 Sentiment Average:   0.65
📊 Fundamental Score:   0.80
🗣️ Analyst Score:      0.55
🔗 Combined Score:      0.67

✅ Recommendation:      Buy

---

Ideal Use Cases

• Day traders seeking instant sentiment snapshots
• Long-term investors validating fundamentals before entry
• Quantitative analysts integrating sentiment into larger models

---

License

This project is licensed under the MIT License. See the LICENSE file for details.

---

Contributing

Pull requests are welcome!

Before submitting, please:

1. Open an issue to discuss your proposed change.
2. Fork the repo and create a feature branch.
3. Write clear commit messages and include tests or examples when applicable.
4. Submit a pull request once your work is ready for review.