AI Upgrade for Pharma Price Audit

I already manage a working price-audit project for pharmaceuticals that sits on 1,875 source PDFs and four associated databases. What I now need is an AI-driven layer that can read those PDFs, tap the existing tables, and produce fully automated audit reports. Key capability required: automated audit reporting that bundles all relevant insights into a single, neatly formatted PDF. Each report must highlight price changes, flag compliance issues, and close with a concise summary of findings. The current assets (PDF corpus + four databases) are well organised and accessible; you will have direct read access. I am happy to expose a small sample first so you can confirm the extraction logic before we go live on the full set. Deliverables • An AI pipeline (Python preferred, but open to alternatives) that ingests the PDFs and database records, then generates the audit report in PDF format. • Customisable report template so I can adjust branding and section order without touching code. • Clear documentation and a short hand-off session showing how to trigger the process and where to tweak thresholds for alerts on price or compliance deviations. Acceptance criteria 1. For any given PDF batch the tool produces a single PDF report containing: – Price change tables – Compliance issue flags with citations – One-page executive summary 2. Processing time stays within five minutes per 100 PDFs on a standard workstation. 3. Source code builds and runs on Windows or Linux with instructions that require no more than 15 steps. If you have experience with NLP on scanned documents (Tesseract, AWS Textract, Azure Form Recognizer, etc.) and can stitch the findings into a polished PDF, this should be straightforward. I’m ready to start as soon as we agree on the approach and timeline.