AI Document Review System

Job ID: 40255207

Budget: ₹1,500 – ₹12,500 INR

I need an internal, AI-driven workflow that lets me upload PDF customer files—primarily bank statements, payslips and KYC documents—and instantly returns a structured report. Once a file is dropped into the system, it should:

• determine eligibility against rule sets I will supply
• extract the salary components and present them in a clean table
• list all obligations in a separate table
• flag any missing or still-required documents
• surface incomplete form fields with exact locations
• generate a short list of probable follow-up queries the credit team is likely to ask

Accuracy on figures and entity recognition is critical, so combine OCR (for scanned PDFs) with NLP techniques; I am open to libraries such as Tesseract, Azure Form Recognizer, Google Vision, spaCy, or a lightweight custom model if that gives better control.

The final deliverable is a web-based or desktop tool (whichever you build faster) that I can host on-prem. Source code, setup instructions and a brief user guide are part of the hand-off, along with a sample output that matches the acceptance criteria above.