AI-Assisted Financial Reporting MVP (2 ERPs → Reports + Presentations)

PROJECT OVERVIEW We are building an MVP of an internal system that automatically collects financial data from TWO ERP systems, consolidates and validates it, and generates reports (Excel/PDF) and PowerPoint presentations. A lightweight AI layer is used ONLY for written commentary and a simple "ask your data" chat — never for calculations. This is a focused MVP, not a full platform. SCOPE (MVP) 1. Data ingestion from 2 ERP systems into one clean, structured dataset 2. Validation / reconciliation and aggregation according to our financial rules (deterministic) 3. Auto-generated Excel and PDF reports 4. Auto-generated PowerPoint presentations from the same data 5. Optional: LLM layer for text summaries + a basic chat to query the consolidated data OUT OF SCOPE (for now) Mobile/desktop apps, multi-tenant infrastructure, multilingual setup, per-record memory. TECH EXPECTATIONS - Python (pandas, openpyxl, python-pptx or similar) - All numeric processing must be deterministic — no LLM-generated numbers - Clean, documented, maintainable code, handed over fully to us DATA SECURITY & LICENSING - Development on ANONYMIZED / dummy data with the same structure; we connect real data at the end - Solution must run locally in our environment, not on your servers - If an external AI API is used, data must NOT be used for training (explain how you ensure this) - Only licensed software and libraries with commercial-use rights - NDA required WHAT WE PROVIDE - Anonymized sample exports from both ERP systems - Field/table mapping and our reporting rules - Templates for the target reports and presentation PROCESS - Start with a small PAID test task (one ERP, one report) before the full contract TO APPLY, PLEASE TELL US - 2–3 examples of similar data-automation / reporting work you've built - Which ERPs you've worked with - How you'd handle data confidentiality - Rough estimate and timeline