Agentic AI Hackathon

Agentic AI Hackathon – National Level: The Agentic AI Hackathon is a 2-day hybrid national-level competition focused on building autonomous, production-ready AI agents capable of reasoning, planning, tool-use, self-correction, and secure execution. Unlike traditional LLM hackathons, this event emphasizes: Agentic reasoning frameworks (ReAct, Reflection, Multi-Agent systems) Tool integration & API orchestration Dynamic replanning under constraints Security guardrails & prompt-injection defense Observability through real-time Thought Trace UI Participants will move from conceptual architecture to a fully autonomous deployed system across three structured rounds. Event Structure: Round 0 – The Blueprint (Online Submission) 5–7 Slide PPT High-Level Architecture Diagram Agent Logic & Tool Strategy Top 40 teams advance to offline rounds. Round 1 – Build the Brain (Offline Day 1) Environment Setup API/Tool Integration First Autonomous Execution Loop Desk-side Technical Review Top 10 teams advance to Round 2. Round 2 – Robustness & Grand Pitch (Offline Day 2) Stress Testing & Edge Case Handling Security Guardrail Audit Final Code Submission 5-Minute Live Demo + Q&A Thought Trace Demonstration Problem Statements Include: Red Team Cyber-Sentry Venture Capital Scout Agent Zero-Waste AI Chef Crisis Response Coordinator Multi-Agent Negotiation System Autonomous Financial Risk Monitor Ethical Decision-Making AI Agent Judging Criteria: Autonomy & Planning – 25% Tool Integration – 20% Security & Best Practices – 20% Robustness – 15% Technical Architecture – 10% Observability & Demo – 10% Event Timeline: Day 0 – Round 0 Submission Deadline Day 1 – Hackathon Kickoff & Development Day 2 – Final Submission & Grand Pitch Eligibility: Undergraduate & Postgraduate Students Team Size: 2–4 members Laptop mandatory Rules: No hardcoded outputs No plagiarism API keys must not be exposed Dynamic reasoning demonstration mandatory Judges' decision will be final Skills to Be Assessed: Artificial Intelligence Large Language Models (LLMs) Agentic AI LangChain LangGraph API Integration Python Multi-Agent Systems Prompt Engineering System Design Cybersecurity Basics Full Stack Development