The Agents' Bounty – AI Agent Build Challenge

The Agents' Bounty is a 3–4 hour individual AI Agent Build Challenge conducted as part of Zettaabyte 5.0. Participants will choose one bounty from a set of practical problem statements and build an AI agent capable of solving it within the given time. The best-performing agents will be evaluated on functionality, AI integration, innovation, UI/UX and presentation, with the top three participants claiming the bounty. Event Details: Date: 11 September 2026. Time: 10:00 AM. Venue: TRCAC, Dahisar East. Duration: 3–4 hours. Entry Fee: Free. Participation: Individual. Format: One-round, offline AI Agent Build Challenge. Problem Statements: 15 total — 5 Easy, 5 Medium and 5 Hard. Bounties: Participants may choose one problem statement. Challenge Categories: Easy — Beginner-friendly AI agent challenges. Medium — Intermediate AI agent challenges involving more complex workflows and functionality. Hard — Advanced AI agent challenges requiring more sophisticated reasoning, workflows or integrations. The difficulty level is intended to help participants choose a challenge suited to their skills. It does not provide any additional scoring advantage. Competition Format: Participants will choose one bounty from the available problem statements at the beginning of the challenge. Each participant will have 3–4 hours to build a working AI agent based on their selected problem statement. The agent must be developed during the event and will be judged and demonstrated on the spot at the venue. Participants may use any AI model, API, framework or service of their choice, including OpenAI, Gemini, Claude, Hugging Face, Ollama and others. Internet access, documentation, libraries and development resources are permitted. Participants may build a web-based UI for their agent where required by the problem statement or their solution. Judging Criteria: All participants will be judged using the same evaluation criteria with a total score of 100, regardless of the difficulty of their chosen bounty. Functionality — How effectively and completely the agent solves the given problem. AI Integration — How meaningfully and effectively AI is used within the solution. Innovation & Problem Solving — Creativity, originality and quality of the approach. UI/UX — Usability, clarity and overall user experience where a user interface is provided. Presentation & Demo — Clarity of explanation and effectiveness of the live demonstration. The top three participants with the highest overall scores will secure first, second and third place. Rules: Participants must compete individually. Each participant may register and participate only once. Participants may select only one problem statement. Collaboration or sharing of solutions between participants is prohibited. Copying another participant's work or submitting a pre-built complete solution is prohibited. The core AI agent must be built during the event. Participants may use any legitimate AI model, API, framework, library or development tool. Internet access and external documentation are permitted. Any malicious, harmful or intentionally disruptive use of AI or other technology is strictly prohibited. Participants must work only on their selected problem statement and their own solution. Any violation of the rules may result in immediate disqualification. The organizers' and judges' decision will be final. Eligibility and Requirements: The challenge is open to college students interested in AI, software development, problem-solving and emerging technologies. Beginners are welcome, and previous experience with AI agents is not mandatory. Participants may use any programming language, AI model, API, framework or development stack they are comfortable with. Participants are expected to create a functional MVP (Minimum Viable Product) of their selected AI agent within the allotted time. The completed MVP will be evaluated and demonstrated locally at the venue as part of the judging process. Rewards: First Place: Certificate. Second Place: Certificate. Third Place: Certificate. Choose your bounty. Build your agent. Prove your creation. Claim the bounty.