Devpost
Participate in our public hackathons
Devpost for Teams
Access your company's private hackathons
Grow your developer ecosystem and promote your platform
Drive innovation, collaboration, and retention within your organization
By use case
Blog
Insights into hackathon planning and participation
Customer stories
Inspiration from peers and other industry leaders
Planning guides
Best practices for planning online and in-person hackathons
Webinars & events
Upcoming events and on-demand recordings
Help desk
Common questions and support documentation
Fold miniature bridges, deliver nine letters and bring a village to life—using your hands.
Make your living room the sound crew. Pair a phone, cast everyday sounds into a twenty-second movie, and premiere your cut on the TV.
Rehearse equipment-lending requests, find missing instructions, and keep committed decisions tied to the rules that allowed them.
Give someone a short-lived pass to ask one question about an object. Review Ring footage privately and share only the answer you approve.
An AI improv game where actors rewrite facts erased from an exact 1,000-token shared memory.
Rehearse a vault withdrawal, inspect the evidence, and keep a local recovery kit that works when the website is gone.
Find the next useful water-use check, then see what changed after a repair.
Review one Python dependency upgrade with advisory evidence, isolated package tests, and a patch you approve before export.
Review a suspicious message, inspect its warning signs, and approve a redacted report to share with someone you trust.
Turn coursework and real availability into a reviewed weekly study plan, then export the calendar you approved.
Turn “two bowls” into a recipe you can use. A CALL E interview clarifies household measures, preserves corrections and ties every scaled quantity to the cook's words.
Calls branches you own, checks their answers against written policy, and reports gaps for review.
One CALL-E task contacts up to five approved suppliers, verifies what each can ship and ranks the evidence side by side in a local review console.
Retired by owner.
A five-minute bomb-defusal game where a human sees the bomb, an AI agent gets seven WebMCP tools, and neither can win alone.
Graft compiles semantic websites into governed WebMCP adapters, then verifies the live tool surface agents receive.
Eleven governed workflow roles collect live cloud evidence, judge SOC 2 controls and ship a hash-verifiable audit package.
One Gemini-fused release-night verdict, human-approved and replayable in ClickHouse.
Private alt-text drafts with rerunnable, on-device Arm benchmarks.
Pre-merge impact analysis for schema changes. Walks DataHub column-level lineage to prove which downstream tables, dashboards and ML assets a column drop breaks, then writes the verdict back.
Family Feud, but the survey is Reddit itself. Guess what the hivemind actually answered on today's hottest thread — 3 strikes, upvote-weighted points, a fresh round every day.
An autonomous software company of 8 AI agents: they bid for work, build via Codex on GPT-5.6 Sol, verify each other with evidence the model can't fake, and settle payment only on proven outcomes.
Long-term memory for Slack: recalls only what fits a token budget, flags contradictions instead of hiding them, and follows you into Claude via MCP.
Vibe working platform using audio and elevenlabs
