Spaces:
Running
Running
| title: Redrob Candidate Ranker | |
| emoji: 🎯 | |
| colorFrom: indigo | |
| colorTo: blue | |
| sdk: gradio | |
| sdk_version: 5.49.1 | |
| app_file: app.py | |
| python_version: "3.11" | |
| pinned: false | |
| # Redrob Candidate Ranker — Sandbox | |
| Live demo of the interpretable multi-stage candidate ranker for the Redrob | |
| "India Runs" challenge (Senior AI Engineer JD). | |
| - Click **Run on bundled 100-candidate sample** to rank a curated slice (strong | |
| fits + keyword stuffers + honeypots) and watch the engine separate them. | |
| - Or upload your own `.jsonl` (≤100 records, schema per `candidate_schema.json`). | |
| Numpy-only at inference time — embeddings are precomputed (`artifacts/`), so no | |
| model runs in the Space. Full code & methodology: | |
| https://github.com/RAK2315/redrob-hackathon | |