Mission
Move beyond chatbot demos and build the trust, coordination, memory, safety, and integration layers that enterprises actually need.
APRIL 10 - JUNE 13, 2026
NandaHack:
Agentic AI
Hackathon
HCLTech · MIT Media Lab
Build a system, tool, or layer that fixes a real bottleneck in how AI agents are deployed, coordinated, or scaled.
What is NandaHack?
A high-signal hackathon for teams building practical infrastructure for enterprise AI agents.
Format
Two phases over ten weeks. Teams prototype, iterate with mentor feedback, and present at a final live event at MIT Media Lab.
Who should join
Builders and engineers
You like shipping real systems, not just demos. Bring your stack and build the missing agent layers.
Researchers and students
Test new ideas in production-like scenarios and get feedback from practitioners, judges, and mentors.
Product and domain teams
Have a real enterprise pain point? Turn it into a solution with measurable business impact.
Why participate
Mentors and feedback
Get direct input from technical and business leaders across enterprise AI and systems design.
Real-world evaluation
Your work is judged on utility, robustness, and how well a stock agent can actually use it.
Career and visibility
Top teams present at the final MIT event and gain visibility with leaders across the ecosystem.
Phases & Tracks
Build solutions that can be used by real enterprise agents in real operational workflows.
PHASE 1Apr 10 - June 13, 2026
Infrastructure Agents
Build agent infrastructure that solves real enterprise problems. On demo day, a judge hands a stock OpenClaw agent your instructions — if it can use what you built, you pass.
| Track | What It Builds | Pipeline Impact | Revenue Lever |
|---|---|---|---|
| Client 0 | Internal AI agents, copilots, productivity tools | Creates proven use cases + credibility | Faster deal conversion (proof-led selling) |
| Enterprise AI (incl. Modernization) | Client-ready solutions (SAP, legacy, agents) | Generates direct qualified opportunities | Large transformation deals ($$$) |
| Sales AI Enablement (Client 0 x Enterprise) | AI-powered account insights, RFP agents | Improves deal identification + win rates | Pipeline expansion + higher conversion |
| Responsible AI | Safety, governance, red-teaming solutions | Reduces enterprise adoption friction | Unlocks regulated industry deals (BFSI, healthcare) |
| Executive AI Coaching | AI copilots for CXOs + coaching model | Opens C-suite access + strategic deals | High-value consulting + transformation programs |
PHASE 2June 13 - July 11, 2026
The Arena
Top teams from Phase 1 enter head-to-head arena challenges. Agents compete live across six arenas — deploy, coordinate, trade, adapt, defend, and integrate — all under pressure.
| Arena | Challenge | Tagline | What It Tests |
|---|---|---|---|
| The Forge | Deploy or die — fastest production-grade agent deployment | Deployment speed & reliability | Infrastructure readiness |
| The Commons | Multi-agent coordination via shared memory | Collective intelligence that works | Coordination efficiency |
| The Bazaar | Agent economy — trade services, verify identity, settle payments | Trust no one, verify everything | Economic infrastructure |
| The Gauntlet | Rapid-fire unknown challenges — pure reasoning under pressure | Adapt or be forgotten | Generalization ability |
| The Wall | Survive adversarial attacks while completing tasks | Last agent standing | Security & resilience |
| The Throne | Full integration — deploy, coordinate, trade, and defend simultaneously | Rule them all | Grand champion |
Judging
The judging is simple and weighted toward substance.
Impact
40%Does it solve a real problem that developers actually have?
Technical Depth
40%Is the implementation solid, thoughtful, and technically credible?
Simplicity
20%Is it easy to understand, adopt, and use?
Timeline
Ten weeks, two phases, one final event at MIT
AnnouncementApr 10
Phase 1Apr 10-June 13
Phase 2June 13 - July 11
Hackathon and Awards at MIT Media LabJuly 11, 2026
FAQ
Who can apply?
Students, builders, researchers, and professionals are welcome. Cross-functional teams are encouraged.
Do I need a full team before applying?
No. You can apply as an individual or as a team. Team matching support will be available before the build phase.
What should we submit?
A working prototype, clear documentation, and a short demo that shows the problem, approach, and impact.
What makes a strong project?
Projects that solve high-friction enterprise bottlenecks with practical, reliable, and testable agent workflows.