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
Every agent is handed tool documentation and told to trust it. This one doesn't. It experiments on its tools, learns what they really do, drops beliefs that stop being true, and publishes the spec.
An agent that closes the loop on its own mistakes, remembering what worked and skipping what didn't, run after run.
Improper payments cost governments and companies hundreds of billions a year. Reckon reads every alert, dismisses the explainable ones, and sends the rest to a named approver with evidence.
An AP agent that flags a price variance once, learns how it was resolved, and never asks about the same case again.
Unwritten learns the rules nobody wrote down—turning past team decisions into an evidence-backed Rulebook that helps agents act with organizational context.
"Don't trust the agent's confidence. Trust what passed."
AI workspace to designing, running, evaluating, and improving specialized agents.
An agent that tries a real GitHub API task, fails honestly, diagnoses its own mistakes, writes itself lessons, and gets measurably better across rounds built entirely inside AO.
The Architect of Efficient Automatic Multi Agent Systems (A-MAS). Built for Track 1. The demo video task shows OCCAM creating an architecture for a well known headache in the FX invoice workflows.
Journeyman is a learning agent that improves at third-party tools without prompt rewrites. It learns from failures, versions its playbook, re-runs safely,and promotes only after hold-out validation.
Give Company Zero an outcome and a budget. It builds the organization needed to deliver it, then restructures itself when execution proves the organization is wrong.
Cascade-Router is a speculative execution router for AI agents that slashes costs by sandboxing cheap student models and seamlessly falling back to powerful teacher models upon schema or logic failure
Engineering, in the loop-track 1
Tacit learns unwritten operational rules from human corrections,applies them only where evidence supports them, and unlearns them when reality changes turning repeated exceptions into safe automation.
AI agents are random, so a score going up might be luck. This system improves agents automatically, but first measures how much the score moves on its own, then rejects any gain smaller than that
Agents that learn the tools, not just the task.
learns how you make decisions from your own apps, no labelling, no training setup. Then it proves what it learned on data it has never seen before.
We break your finance agent on purpose - then teach it not to break the same way twice. MAYDAY RECON: because production is a terrible place to discover edge cases.
A self-improving AI agent that learns how to use unfamiliar tools.
An autonomous agent engineering system that automatically designs, executes, benchmarks, diagnoses, and improves specialized AI agents through closed-loop hill climbing.
MORPHOS is an autonomous agent-engineering platform that evolves AI agents through real execution, failure diagnosis, targeted mutation, memory, and multi-objective evaluation.
An agent that learns a codebase's unwritten conventions by writing down what it learns, plus an outer loop that rewrites the agent's own code when memory isn't enough.
a narrow autonomous AR agent that learns a short-pay policy, runs it against incoming cash, escalates exceptions with accountant-legible reasoning, and improves after repeated human feedback.
AI-Powered Finance Workflows, Built for the Office of the CFO
1 – 24 of 101
