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
Sort by:
An open, model-agnostic, multi-track asynchronous reasoning harness and heuristic governor optimized for high-concurrency, low-latency execution on Arm64 architecture.
Secure NetOps for Distributed Infrastructure, aka: New Technology IP
A Non-Predictive, Correction-Aware Monitor It shows what entered an information field, what was clustered, what was admitted, what was held out, what changed, and what remains unresolved.
Teacher Brain supports underserved communities facing teacher shortages. It recognizes raised hands, answers in each student’s language, and turns questions into engaging visual lessons.
Practice high-stakes negotiations against a resistant AI counterpart that won't fold to weak arguments — then get scored coaching on your anchors, concessions, and tactics.
The live MEES compliance command centre for UK letting agents — portfolios matched to the official government EPC register, scored into a risk queue, and actioned before a property blocks a letting
Stillwater turns routed MoE experts into a storage-backed execution backend for Hugging Face Transformers, with exact native parity and configurable memory residency across multiple model families.
The project is is assist the developer to controll swarms of agents and the track token consumed and cost , it is a planning and llm ocherstrator.
🚀Project Northline: A zero-telemetry AI enterprise. Powered by GPT-5.6 & Codex, our autonomous Chief of Staff directs 15 offline HR agents, ensuring absolute IP security and corporate continuity.
An AI-powered evaluator: stop eyeballing RAG outputs - GPT-5.6 scores, diagnoses, and auto-tunes your pipeline automatically.
One QR Code, everything from your talk
Upload any repo. Get AI-generated READMEs, architecture diagrams, and API docs in seconds — no more stale documentation.
An evidence-first trading platform that turns market news into transparent, risk-controlled options decisions and trades.
UniVerse unifies campus tools with AI, organizing everything so students focus on learning. "Put Your Studies in Orbit"
It knows what you want: A local-first recipe that learns how you decide, predicts feedback, and escalates instead of guessing.
A mobile Windows command center for Claude, Codex, OpenCode, and ACP agents working across your Obsidian vault and projects, with persistent chats, visible tool use, approvals, and usage analytics.
An AI that doesn't grade wrong answers — it diagnoses the belief that caused them, then fixes it for good.
WriteAIBook turns one premise into an editable manuscript that remembers its characters, plot, and voice across every chapter—powered by GPT‑5.6 Terra.
AI safety check for medical scan report drafts to avoid mistakes and pitfalls by Doctors
A shared hiring memory inside Codex - turning CVs, emails, and evidence into clear action, then using the Codex plugin ecosystem to draft outreach, schedule calls, and move hiring forward.
Making human judgment inspectable—from source and AI interpretation to coherent action. Perception, recognition and coherence as a strategic asset for the human layer.
A local-first AI command center built with Codex and GPT-5.6 that unifies scattered files, AI outputs, and workflows into one visible, searchable, action-ready workspace for business owners.
Jupiter makes computers accessible for people with limited mobility. By simply expressing what they want to do, users leverage ChatGPT to control their PC effortlessly
Runtime governance for AI agents: one authenticated decision authorizes one exact action, for at most one dispatch attempt, with auditable policy and human review.
