Loops House


Metro Saathi
Multilingual complaint desk for Pune Metro Aqua Line. AI understands the report, rules decide the priority.

DevDay Exchange Community Metro Hackathon: Pune
Team
1 member- ISOwner
Ishan
Overview
Metro Saathi is a multilingual complaint desk for Pune Metro's Aqua Line.
Commuters report a problem in Marathi, Hindi or English (or Marathi typed in English letters). OpenAI turns the messy text into clean signals — category, danger, vulnerable commuter, crowding — and our own rule-based formula sets a 0–100 priority with a readable reason, so every score can be explained. If the AI is unavailable, keyword matching takes over and the system keeps working.
Staff get an Issue Desk: emergencies pinned at the top, reports grouped by station, an Aqua Line map showing where problems are, and related reports at neighbouring stations within 60 minutes grouped into one incident. Each report moves open → in progress → resolved, with technician notes. Commuters track progress with a reference ID like MS-1043.
Built in one day with OpenAI Codex. 12 automated tests. Live, deployed and working.
Live: https://metro-saathi.onrender.com (staff: https://metro-saathi.onrender.com/staff) Repo: https://github.com/Ishan-246/metro-saathi