Case

Next.js · TypeScript · SQLite · OpenAI API

PCSmilesAI

A multi-location dental group was spending hours every week opening invoice PDFs, guessing vendors, and assigning costs to clinics.

Constraint

Invoices were living in a dental inbox. Staff hunted through email, named vendors by hand, and typed the same numbers into spreadsheets — across multiple locations.

What we did

I built a pipeline that takes invoices in by email, reads them with AI, matches vendors, maps costs to the right clinic, and puts everything in a searchable dashboard. People review the edge cases. The rest runs.

Result

Hours of admin work left the week. The team stopped treating the inbox like an accounts-payable system.

Braxton's work has saved us countless man hours in our admin processes and has really helped us understand how to leverage AI in a much more meaningful way.
McKay Marshall · PC Smiles