Spreadsheet-to-Database AI Migration

I currently run an in-house spreadsheet application that has outgrown its limits. I want to migrate everything into a purpose-built database that can accept multiple incoming feeds—CSV files arriving by batch, live API calls, and occasional manual entries (often straight from Excel).

Before any record lands in the tables it needs to pass through an AI layer that cleans, classifies, and enriches the data. Once stored, different user groups must see only what their privilege level allows, so role-based security is essential.

My short-term goal is to generate a print-ready publication from the data; longer term I will also push the same content to a website and, potentially, a mobile app.

Deliverables
• A relational schema that mirrors the existing spreadsheet logic while remaining flexible for future fields
• ETL/ingestion scripts or workflows that import CSV, pull from APIs, and accommodate keyed-in updates, all routed through the AI preprocessing step
• An administrative interface where I can define and adjust user roles and permissions
• Export routines that assemble the data into a printable layout today and can be repurposed for web or app delivery tomorrow
• Clear deployment instructions (preferably Docker-based) and a short hand-off session so my team understands day-to-day operation

I’m open to the stack—PostgreSQL, MySQL, or even a managed cloud database—as long as it stays cost-effective and scales gracefully. Python, Node.js, or similar languages are fine for the ingestion and AI logic; use whatever you’re most productive with.

If this sounds like your wheelhouse, outline briefly how you would structure the build and how long you expect each milestone to take.