MaintenanceTracker
About the project
It is an open-source, privacy-first equipment maintenance tutor and inventory scheduler designed for research, university, and industrial laboratories to manage hardware uptime without exposing sensitive facility data to the cloud. Built on a React + Vite frontend and a Python FastAPI backend, it leverages Gemma 4 running locally via Ollama to visually inspect damaged hardware from photos, auto-onboard equipment from model labels, perform Q&A across multi-page PDF repair manuals, and manage calibration schedules or import legacy CSV/Excel inventories—all executed entirely on-device.
Built with
gemma
AI tools
Coded withClaude CodeCodexGoogle AI StudioOpenCode
Prize Categories Entered
- Submitted
Best Use of Gemma 4
- Submitted

Best Open-Source AI Project