⚖️ Legal Case Matcher Based on Client Name with Gemini AI Summarizer

A Python-based tool that matches legal cases by client name using semantic search, and simplifies case summaries using Gemini (Google AI). Case data is sourced from a .docx file.

📂Project Features

🔍 Client Name Matching using Sentence Transformers + Fuzzy Matching
📄 Reads Case Data from a Word .docx file
🤖 Uses Gemini API to Simplify Legal Summaries
🔐 API key securely managed via .env file

✅Requirements

• Python 3.8+
• Install dependencies via:

pip install -r requirements.txt

▶️How to Run

1. Place your cases.docx file in the same directory as the script.
2. Create a .env file in the root directory and add your Gemini API key:

GEMINI_API_KEY=your_gemini_api_key_here

3. Run the script:

python lawfirm_case_matcher.py

4. Enter the client’s name (full or partial) when prompted.

---

🔍What the Program Does

• Matches the closest case from your .docx using:

  • SentenceTransformer-based semantic similarity
  • Fuzzy string matching for partial names

• Displays:

  • Client Name
  • Charges
  • Case Summary

• Sends the case summary to Gemini for simplification, and displays a human-readable explanation.

📝DOCX Format Example

Neha v. Suresh (2023)
Summary: Dowry harassment post marriage.
Section: IPC §498A, §406

State v. Meena Kapoor (2020)
Facts: Meena allegedly set her in-laws’ house on fire during a domestic dispute.
Section: IPC §436, §498A

• Separate each case by a line.
• "Summary" or "Facts" line must include a colon (:).
• "Section" entries should be comma-separated.

🧠Gemini Output Example

Enter the client's name: neha

Closest Match Found:
Name     : Neha v. Suresh (2023)
Charges  : IPC §498A, §406 (Criminal breach of trust)
Summary  : Dowry harassment post marriage.

Gemini Summary (Simplified):
Neha reported harassment from her in-laws over dowry after marriage. Legal charges were filed under IPC sections §498A and §406.

Notes

• Ensure your Gemini API key is correct and active (free-tier limits apply).
• If you're rate-limited or over quota, the Gemini output may be blank or show an error.

License This project is free and open-source for learning and legal research use cases.

Credits

• Google AI Gemini API
• Sentence Transformers
• RapidFuzz
• python-docx

👩‍💻 Built by: xoxo