Knowledge validation
Initial quiz-based evaluation and participant selection.
- → 45 minutes, online, individually attempted
- → ML fundamentals, Python, model lifecycle, ML Ops tooling
- → Shortlist published within 48 hours
AI Engineering & ML Ops Hackathon Builder
█
A progressive, university-focused AI engineering and ML Ops buildathon designed to move participants from knowledge validation to MVP development and an intensive on-premises build experience.
Hosted by the IEEE Computer Society of KDU and open to undergraduates across Sri Lanka. Three phases, one shortlist, and a final day that ends with your project on a stage.
Initial quiz-based evaluation and participant selection.
Selected participants receive an organizing-committee-defined problem statement and build an MVP.
Six-hour on-premises buildathon followed by project pitching, evaluation, awards/recognition and the closing ceremony.
Full rulebook published with the announcement on 12 Sep 2026.
Phase 3 pitches run seven minutes plus three for questions. Every project is scored against the same five axes, weighted as below. Provisional weights — confirmed with the rulebook.
Cash prize (value TBC), the TensorForge trophy, and an introduction to the knowledge partners' engineering teams.
Cash prize (value TBC) and partner-sponsored cloud or compute credits for the team.
Cash prize (value TBC) and course or certification vouchers for each member.
For the team whose deployment, monitoring and reproducibility story is the most complete.
For the strongest first-year or first-time team on the floor.
Voted on the day by everyone in the room, participants included.
Discussions and confirmation are ongoing.
Interested in partnering? Talk to the committee →
Drag photographs onto any frame to fill it.
Names to be confirmed
Anything still unclear, mail the committee and we will add it here.
No. Register as an individual and attempt Phase 1 alone. Teams are formed after the shortlist, and the committee helps solo participants find one.
Machine-learning fundamentals, Python, the model lifecycle and basic ML Ops tooling — containers, versioning, deployment and monitoring concepts. A study outline goes out with the announcement.
No. Entry is free for all three phases. Refreshments on the final day are covered by the organisers.
At KDU on 10 October 2026. The exact hall and reporting time are sent to shortlisted teams with their Phase 3 confirmation.
Your own laptop and charger, your student ID, and any hardware your project needs. Power, network and workspace are provided.
Yes, and we expect you to. Declare what you used in your pitch — judges care about the engineering decisions around the tools, not abstinence from them.
Yes. One statement, defined by the organizing committee, released to every shortlisted team at the same moment.
Everyone who competes on the final day receives a certificate of participation. Winners and special-recognition teams receive theirs at the closing ceremony.
Registration runs 14–22 September 2026. One entry per person; you will get the Phase 1 quiz link by email on 26 September.