What is included:
• Multi-format ingestion (scanned PDFs, bank statements, tax forms, images)
• High-accuracy OCR & vision extraction with automated table and signature parsing
• Retrieval-Augmented Generation (RAG) with vector databases (Pinecone, Qdrant, pgvector)
• Human-in-the-loop review interface for low-confidence data flags
• RESTful Python (FastAPI) microservice ready for cloud deployment