Legiscan

About the Project

Inspiration

Manual contract reviews are time-consuming, error-prone, and costly. This inspired me to build an AI-powered tool that simplifies and automates contract analysis, ensuring compliance, reducing risks, and saving time for businesses and legal professionals.

What I Learned

I gained insights into embedding generation, similarity search, and integrating AI with legal workflows. I also enhanced my skills in database management, Docker, and report generation.

How I Built It

• Used Gemini LLM to generate embeddings for contract clauses.
  
• Stored embeddings in a PostgreSQL database hosted via Docker.
  
• Performed similarity search to analyze compliance.
  
• Generated a detailed report with compliance scores, strengths, weaknesses, and recommendations.
  

Challenges Faced

• Extracting clean text from varied file formats.
  
• Fine-tuning similarity search for precise clause analysis.
  
• Managing large embeddings efficiently in a database.
  

Built With

• docker
• gemini
• genai
• langchain
• postgresql
• python
• tensorflow