Legal Document Review Assistant – Personal Project
Designed and deployed an AI-powered legal document analysis system that automates contract review by extracting key clauses, identifying risks, and summarizing terms using Llama-3 and GPT-4. Integrated OCR pipelines, FAISS-based vector retrieval, and multi-agent reasoning via AutoGen, enabling clause verification and cross-document comparison. Reduced manual review time by over 60% for legal teams while improving clause-level consistency and compliance tracking. • Developed prompt engineering workflows for contract clause extraction. • Conducted document-level text summarization and risk identification through LLM prompting. • Utilized OCR for scanned document digitization and classification. • Employed multi-agent LLM architectures for clause verification and cross-comparison.