SWHSim Wen Hao
Email

Sales, purchase and inventory planning app

Built and verified Sep 2026
  • Flask
  • SQLite
  • React
  • Playwright

Planning for one business unit ran on three linked Excel workbooks and 95 queries, refreshed by hand in a fixed order. I built a web application that rebuilds the same figures on a schedule.

37/37review groups matching the workbooks
Real interface, sample data. Rows are tinted by months of cover: running out, too much stock or not selling.

Method and results

  1. It rebuilds all 76,028 monthly cells from their order lines and matches the workbooks on 37 of 37 review groups.
  2. Quantities are keyed by the month's date, after a month once shifted position in the source files.
  3. The stack is Flask with SQLite, a React front end, company single sign-on and a Python ingest that reads the .xlsm files directly.
  4. API calls took 2,031 ms until I found that Windows was resolving "localhost" to IPv6 first. After the fix they took 3 ms, and the main payload went from 420 KB to 36 KB.
  5. It has 182 browser checks and 10 server check scripts.