Agent-a-thon is a 24-hour engineering hackathon where participants design and deploy multi-agent AI systems that can plan, act, and make autonomous decisions. (Open for all B.Tech branches.) Build Autonomous AI Agents for a Better Tomorrow
Rules & Guidelines:
Teams: 2-3 members
Build everything during the event.
No pre-built or plagiarised solutions.
Time: 24 Hours
Presentation:
8–10 min per team
Must follow official submission format provided during the event.
Theme: Engineering Autonomous Agents for Nature & Safety
Track 1: AI for Nature:
Build Agents For: Pollution / Air Quality | Wildlife Alerts | Water Quality | Campus Green Index | Waste Detection | Smart Irrigation
Track 2: AI for Safety:
Build Agents For: Scam Detection | Cyber Safety | Campus Incident | Reporting | Women’s Safety | Hazard Detection | Risk Evaluation | Build Requirements
Teams Must Create a System With:
Min. 2 Autonomous Agents
A Defined Workflow (planner Executor Evaluator Workflow)
Safety Boundaries
Demo-Ready Ui (web/app/cli/simulation)
Allowed Frameworks: Open Ai Assistants, Lang Graph, Crew Ai, Auto Gen