# Warriorme

- **Event:** [Meta Global AI Developer Hackathon](https://cerebralvalley.ai/e/meta-ai-developer-hackathon)
- **When:** Oct 26, 12:00 PM – Nov 4, 9:00 PM (EST)
- **Where:** Online
- **Team:** [Devanandu Sreenath](https://cerebralvalley.ai/u/deva03)
- **GitHub:** https://github.com/Devanandusreenath/route_autopsy
- **Demo video:** https://drive.google.com/file/d/1ku4reJhK0sPfxWygV0hIXyJEPjag9bQB/view?usp=sharing
- **Gallery:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2157

Route Autopsy — Don't Just Trust a Route. Investigate How It Could Be Wrong.

Traditional navigation systems recommend the fastest or shortest route, but they rarely explain how reliable that recommendation is when road conditions, vehicle restrictions, or uncertain information change.

Route Autopsy is an AI-powered route investigation and decision intelligence platform that evaluates routing recommendations rather than blindly trusting them.

It combines intelligent route calculation, vehicle-specific restrictions, road defect analysis, counterfactual simulations, and Meta multimodal AI to investigate the reliability of a route.

Users can calculate routes, explore alternative paths, and run what-if scenarios such as road closures, severe potholes, new restrictions, or changes in vehicle specifications. The Autopsy engine identifies which assumptions could change the recommendation and explains why the route changes.

Using Meta multimodal AI, Route Autopsy can interpret road images and extract candidate evidence about defects and road signs. These observations are presented for human review before becoming verified road intelligence.

Our key innovation is counterfactual route investigation: instead of simply asking which route is best, we investigate what could make that recommendation wrong.

Built using Python, FastAPI, React, TypeScript, and Meta multimodal AI, Route Autopsy aims to make mobility decisions more transparent, evidence-driven, and trustworthy.

## More from Meta Global AI Developer Hackathon

- [hadiproz](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2154)
- [Emmanuel Onwuka](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2155)
- [ProofMerge](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2156)
- [PrivLab](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2158)
- [JustNewMe](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2159)
- [FundiLoop](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2160)

---

Markdown version of https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2157. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt