Built a document processing system using multimodal AI for a legal tech startup. The platform extracts key information from contracts, invoices, and legal documents — processing images, PDFs, and scanned documents with 95% accuracy, cutting manual review time by 80%.
Technologies: Python, FastAPI, AWS
Results: 95% extraction accuracy, 10x faster document processing