A NEW AGENT - Hackathon
- Date
- 2025-10-11
- Location
- San Francisco, CA, USA
- Host
- AI Valley
About this event
A NEW AGENT - Hackathon is a full-day, in-person build sprint for people who want to move from talking about AI agents to actually creating them. If you’ve been waiting for the right room, the right momentum, and the right level of technical energy to ship something real, this is that day. About the Event This hackathon is designed around one central idea: agents are becoming a practical interface for software, and the best way to understand that shift is to build. Rather than a passive conference format, this is a hands-on day for making, testing, collaborating, and learning in real time with other people who care about AI, product, and engineering. You can come with a rough concept, a half-formed problem statement, or simply curiosity about what a strong agent workflow looks like in practice. The format supports both people who already know what they want to build and people who want to find teammates, explore ideas, and get moving once they arrive. Expect a community that is technical, ambitious, and open. The point is not polished demos for their own sake; it’s to create the conditions for serious experimentation, useful feedback, and fast iteration over the course of a single day. What to Expect The day begins Saturday, October 11 at 9:00 AM PDT in San Francisco with the kind of kickoff that gets everyone oriented quickly and into build mode. Early on, attendees can expect time to settle in, meet other participants, surface ideas, and identify the kinds of agent projects they want to pursue. From there, the core of the event is the hackathon itself: focused work sessions where teams and solo builders can design workflows, test models, wire up tools, and figure out what actually works. This is the part of the day where concepts get pressure-tested against real implementation constraints, which is often where the best learning happens. Throughout the event, you should expect a mix of: Idea sharing and team formation for people looking to collaborate Dedicated build time to develop and refine an agent-based project Peer feedback and conversation with others working through similar technical questions Show-and-tell energy that helps turn experiments into something others can react to Because this is an in-person event, a lot of the value will come from the room itself. You’ll be able to compare approaches, troubleshoot faster, and make progress with people who understand the same technical and product challenges you’re working through. Why Attend If you’re interested in AI, there’s a big difference between reading about agents and trying to build one under real constraints. This event gives you a clear, practical environment to do exactly that. You’ll leave with sharper instincts about tooling, orchestration, prompting, interfaces, and where agent ideas become genuinely useful versus unnecessarily complicated. It’s also a good setting to find your people. Hackathons tend to attract participants who are willing to experiment, move quickly, and share what they’re learning as they go. If you’re looking for collaborators, future teammates, or simply a stronger local network around AI and technical building, an in-person day like this creates much better conversations than a typical online thread or content feed. You should also expect the value of compressed time. Spending a focused day in build mode can accelerate a project that has been sitting in your notes for weeks. Even if you don’t arrive with a finished concept, you can still leave with a prototype direction, a clearer problem framing, and a stronger sense of what to build next. For founders, engineers, designers, researchers, and technically curious builders, the biggest takeaway may be clarity. What kinds of agent experiences are actually compelling? What infrastructure matters? Where do workflows break? Which ideas deserve another week of work? This event is a fast way to get better answers. Practical Details This is an in-person event in San Francisco, USA, which means you should plan for a live, collaborative environment with all the advantages that come from building alongside other participants. Face-to-face time matters at a hackathon: it makes team formation easier, feedback faster, and problem-solving much more fluid. The event starts Saturday, October 11 at 9:00 AM PDT. Because this is a hackathon, it’s smart to arrive ready to work: bring your laptop, chargers, any tools or accounts you expect to use, and at least one idea you’d be excited to explore. If you don’t have a full concept yet, that’s fine too; showing up with curiosity and a willingness to collaborate is enough to get started. A few useful ways to prepare: Think about a problem area you care about before you arrive Be ready to explain your idea simply if you want to find teammates Come open to changing direction once you see what others are building Plan for a full, focused day rather than a casual drop-in If you want a day that is practical, social, and deeply tied to the future of AI software, A NEW AGENT - Hackathon is built for that purpose. Show up ready to make something.
Who should attend
This is for you if you want to spend a day building around AI agents with other motivated people instead of learning in isolation. - **You’re an engineer or technical builder** who wants a concrete setting to prototype agent workflows, test tools, and see what holds up in practice. - **You’re a founder or product-minded operator** exploring how agent-based experiences could create real user value, not just novelty. - **You’re a designer or researcher** interested in how people will interact with intelligent systems and want to shape those experiences through hands-on experimentation. - **You’ve been sitting on an AI idea** and need a deadline, a room, and a community to finally start building it. - **You want collaborators** and know that the fastest way to find strong technical peers is to work beside them on a real project. - **You’re curious about the current state of agent development** and learn best by shipping, testing, and getting feedback rather than just reading or watching demos. You do not need to arrive with a polished concept. If you’re excited by AI, comfortable with fast-moving technical conversations, and ready to contribute, you’ll fit right in.