Hi everyone,
I recently open-sourced a project called JobMatch Bot.
It’s a Python pipeline that aggregates jobs directly from ATS systems such as Workday, Greenhouse, Lever, and others, normalizes the data, removes duplicates, and ranks jobs based on candidate-fit signals.
The motivation was that many relevant roles are scattered across different company career portals and often hidden behind filtering mechanisms on traditional job sites.
This project experiments with a recall-first ingestion approach followed by ranking.
Current features:
• Multi-source ATS ingestion
• Job normalization and deduplication
• Resume-aware ranking signals
• CSV and Markdown output for reviewing matches
• Diagnostics for debugging sources
It’s still an early experiment and not fully complete yet, but I wanted to share it with the Python community and get feedback.
GitHub:
https://github.com/thalaai/jobmatch-bot
Would appreciate any suggestions or ideas on improving ATS coverage or ranking logic.
[link] [comments]







English (US) ·