In this this 10+ hour intensive hackathon led by the research team at [Obvious](https://obvious.ai/frontier) you will build your first Level 2 autonomous software factory. TLDR: you'll be able to start assigning your agents features to build instead of prompts.
REGISTER ON LUMA: [https://luma.com/frontierATX](https://luma.com/frontierATX)
*Note: This event is free to attend with the exception of the tokens you use during the hackathon. Some work will require an [obvious.ai](https://obvious.ai) workspace (subscription fees waived, inference or compute used will be passed on to you).*
Come alone or come as a team. If you want to valuemaxx this event the ideal group to bring is:
• a technical founder or engineering leader
• a product manager (if you have one)
• 1+ engineers (the more ai-pilled the better)
What to bring
1. Your laptop
2. Access to a real Github repo you care about [or] a clear vision for a greenfield project
3. A bug backlog or a list of small improvements
4. An Obvious workspace. Sign up and follow the onboarding steps (will be shared in advance).
Meals and snacks are provided. 🍕🍻
What we'll be covering in the sessions:
How to...
• design the complete hands-off build loop (coding agent of your choice)
• manage a multi-pr autonomous agent swarm
• prepare more ambitious builds (entire features or even products)
• manage QA, code review, and human feedback without slowing down the work
• decide where to spend time vs go hands off
• rethink both devops and organizational roles in a world of autonomy
• scale autonomous software delivery safely
• think about roadmap and software design when you can build anything practical in a day
What you'll be working on:
Level 1
• designing the core verification, monitoring, and "ralph" steps of your loop (how you trust the outcomes)
• context engineering improvements to allow agents to operate without constant steering
Level 2
• executable sandbox environments in the cloud (give the agents a computer)
• optimizing CI to be compatible with a rapidly looping factory
• setting up agentic reviewers (if you don't have them)
• determining autonomous inputs to your factory (bug reports, feature requests, etc)
• connecting your factory to your sources of work, feedback, and context
-------
About Obvious:
Obvious is an applied AI research lab with a focus on autonomous product development. We spend our days closing the gap between what's possible and what's doable, treating our own company as the experiment for what it means to operate at agent scale. We've reinvented our product, reshaped our team, accelerated revenue, and 10x'd our shipping velocity in less than a year. At peak velocity we've shipped 1000 PRs in one day with 10 engineers.
Frontier is where we do our best to hand that system to other teams. We're excited to work with you!
→ Learn more at[ obvious.ai/events](https://obvious.ai/events)
