Our mission is narrow and hard: close the distance between building an AI demo and shipping an AI system.

About

Most programmes reward the pitch. CUBE rewards the engineering behind it: what you built, why you built it that way, and what happened when it met somebody else’s code.

3

Rounds

Individual build, pod integration, live defence

5

Challenge domains

Real intelligent-commerce problems

₹50,000

Cash prize pool

Awarded at the Grand Finale

Evidence over information

We do not score how much you built. We score what you can prove, explain and defend: your repository, your reasoning, your answers under pressure.

Integration over isolation

Most programmes end where the hard part begins. Round 3 starts the moment your agent has to depend on four agents it did not write.

Engineering over administration

Your weekend belongs to reasoning loops and failure modes, not scaffolding. The starter repository already runs, so the first hour goes into the actual problem.

Transparency over black boxes

Every judgement comes with its reasoning. You learn why your system placed where it did, and you are asked to do exactly the same for it at the defence.

Sydon.AI
Writes the problem statements, reviews the code, sits on the defence panel.
CodeQuesters
Brings the builders, runs the kickoff, and forms the pods.
Practising engineers
Every submission is reviewed inside its own domain, by people who solve that problem for a living.
₹50,000 prize pool
Awarded at the Grand Finale on 10 October 2026, alongside recognition and a genuine career conversation with the Sydon.AI engineering team.