Fairmarkit
Hackathon
We're picking the boldest builders at Fairmarkit. Eight tracks, one week, one shot to ship something extraordinary.
Hacking starts in
Wed, May 6 · 8:00 AM Warsaw
Schedule
1-hour kickoff & Q&A. Teams submitted by 6PM EST. Assignments by 7PM.
8AM Warsaw. Full hacking window begins. Story points reduced.
8–10AM EST daily. Book 15–30 min with VK and AT.
Teams report progress. 8AM–12PM EST. Book 15–30 min.
All teams present. Top 3 advance to finals.
Top 3 re-demo. 30-min ceremony. Prizes announced live.
Rules & Structure
Teams
1–3 people per team. Solo allowed. 3 is the hard cap. 3–4 teams max per agent track.
Participation
Product & Engineering mandatory. CS / Sales / Marketing voluntary. Sprint points reduced.
Timeline
Submit team + agent preferences by 6PM EST Tue. Assignments released by 7PM. Hacking starts Wed 8AM Warsaw.
Help Committee
AT and VK available for org questions, infra support, AI guidance, business opinions, and technical challenges.
Scoring
All tracks: 11 general pts + up to 6 bonus pts = 17 pts max. Jury: Kevin, AT, VK.
General Criteria (max 11 pts)
Each agent track also has up to +6 bonus points for track-specific criteria.
View all track details & bonus criteria →Tech & Patterns
Recommended Stack
Recommended but not limited to — use whatever makes you faster, more advanced, and production-ready.
Graph-based agent orchestration with full observability
Type-safe agent framework with structured outputs
LangSmith Studio is highly recommended for complex agentic flow debugging — trace every step, inspect tool calls, and replay failures.
A browser plugin is a great option to integrate your agent with the platform and demo on real data within the real environment.
Both direct API and MCP integration are in scope — choose whichever fits your agent best.
Agentic Patterns
These patterns score bonus points. Explore the interactive diagrams below — and go beyond.
Suggested Day-by-Day Approach
Generate concepts, split ideas for evaluation, and relentlessly vibe-code to validate all ideas fast. Ship throwaway POCs — speed over polish.
Evaluate all ideas and vibe-coded POCs with knowledge keepers: Sales, CS, AT, and VK. Highly advised. The more feedback you collect now, the fewer surprises on evaluation day.
Integrated, stable, UX-polished. Mix of AI-assisted and manual coding depending on your skills. Focus on the complete flow, not just the happy path.
Last in-person feedback opportunities, cleanup, fixes, evals, integration testing, and evening presentation to all teams.
Top 3 re-demo for full audience. Jury announces prizes live. 30-minute ceremony.
Ready to build?
Claim your spot, form a team, pick your agent track, and start hacking.
Claim Your Spot