Our back-office is drowning in invoices, contracts, and employee records that arrive in every imaginable format. I want a single AI-powered system that takes each document as it comes in, recognizes what it is, and files it where it belongs—ready to be found again in seconds. Here’s what the platform has to do for me: • Automatically classify each incoming file as an invoice, a contract, or an employee record, even when the layout varies. • Maintain clean version control so I can track every change and roll back if needed. • Provide lightning-fast search and retrieval by keywords, dates, or any tagged field, with OCR handling the scanned paperwork we still receive. I’m comfortable if you build on Python, Node, or any modern stack, as long as the core engine uses dependable AI/ML libraries (e.g., TensorFlow, PyTorch, spaCy) and an indexed search layer like Elasticsearch or similar. A lightweight web dashboard or a set of well-documented APIs is fine; the key is accuracy and speed. Deliverables I expect: 1. A working prototype deployed to a cloud test environment. 2. Source code with clear setup instructions. 3. A concise user guide demonstrating categorization, version control, and search on sample invoices, contracts, and HR records. If you have tackled document heavy workflows before—especially with OCR and intelligent tagging—let’s talk.