Threat Evaluation Report – AI Safety Analysis
Built at the Robotic Agents Hackathon. An AI assistant that analyzes images and text for physical threats such as fire, smoke, structural damage, and electrical hazards, then returns a structured assessment with severity, urgency, and recommended actions. Evidence-gathering tools pull evacuation routes and building sitemaps to back up the report. The backend was designed to accept video from unmanned ground vehicles through Cyberwave, so a patrol robot could feed the same pipeline.
Project Features
Image and Text Threat Analysis
Upload a photo or describe a situation and get a structured report instead of a paragraph: hazard, severity, urgency, actions.
Evidence Tools
Smart Search for evacuation routes and sitemap extraction tied to a specific building, so recommendations reference real exits.
Robot-Ready Endpoints
POST endpoints for video frames were designed in from the start so a UGV feed could replace manual uploads.
Chat Interface
Image uploads, note annotations, and tool results side by side in a two-panel React UI.
Project Info
03/2026
Backend & Agent Tools
Hackathon Team
Hackathon Submission
Technologies Used
- • Python, FastAPI, LangChain
- • Featherless AI (Qwen)
- • ScrapeGraph
- • Smallest.ai
- • Cyberwave
- • React, Vite
Project Gallery
Interested in This Project?
I'm always excited to discuss my projects and share insights about the development process. Feel free to reach out if you'd like to know more!