OSS Agents: Fine-Tuning to Production
- Date
- 2025-10-23
- Location
- San Francisco, CA, USA
- Host
- Runpod
About this event
Open-source agents are moving fast, but the gap between a promising demo and something you can reliably ship is still where most teams get stuck. OSS Agents: Fine-Tuning to Production is for builders who want a clearer path from experimentation to real-world deployment, with practical conversations about what actually matters once an agent has to perform outside a notebook. This in-person San Francisco meetup brings together people working at the intersection of AI, open source, and production systems. Expect a room full of engineers, founders, researchers, and curious practitioners comparing notes on how to make agent workflows more capable, more reliable, and more useful. About the Event This event is built around a simple idea: open-source agents are only interesting if they can survive contact with real users, real data, and real constraints. The focus here is not abstract hype. It is the messy, technical, and often very interesting work of taking agent systems from fine-tuning and evaluation through deployment, iteration, and ongoing improvement. The format is designed to be social and practical. Rather than a purely one-way presentation, this meetup is meant to create room for both structured discussion and informal conversation. You should come expecting a blend of focused insights, shared lessons from the field, and the kind of side conversations that often end up being the most valuable part of the night. Because this is an in-person community gathering, the event also serves as a strong networking moment for anyone building with OSS AI tooling in the Bay Area. If you have been looking for peers who care about agent reliability, orchestration, fine-tuning decisions, evaluation setup, and production tradeoffs, this is the kind of room where those conversations happen naturally. What to Expect The evening will center on the practical lifecycle of OSS agents, from model adaptation through production readiness. While the exact flow may vary, the event is likely to move through key themes such as how teams are approaching fine-tuning, what changes once agents are exposed to real user behavior, and which operational challenges show up fastest in production. You can expect discussion around topics like: Fine-tuning strategy: when it helps, when it is unnecessary, and how teams decide where to invest effort Agent architecture: how people are structuring workflows, tools, memory, retrieval, and orchestration in open-source stacks Evaluation and reliability: how to think about testing agent behavior before and after launch Production lessons: latency, cost, observability, safety, failure modes, and iteration loops Open-source ecosystem insights: what practitioners are using, adapting, and learning from right now There will also be plenty of room for direct peer exchange. That means you should expect conversations that go beyond polished answers: what broke, what surprised people, what worked better than expected, and where teams are still uncertain. For many attendees, that level of honesty is what makes a meetup like this useful. If you are actively building, this is a good place to bring your own questions. If you are earlier in your journey, it is a strong opportunity to listen for patterns and get a more grounded understanding of the path from prototype to production. Why Attend If you work with AI systems, you already know that building an agent that looks impressive in a controlled setting is very different from building one that is dependable in practice. This event is valuable because it focuses on the transition point where architectural choices, tuning decisions, tooling, and operational discipline start to matter together. You will leave with a better sense of how other builders are thinking about the full stack around OSS agents. That can help you sharpen your own roadmap, avoid avoidable mistakes, and spot where your current setup may need stronger evaluation, better observability, or a different approach to model adaptation. There is also strong value in the community itself. San Francisco is one of the best places in the world to meet people working deeply on AI infrastructure and applications, and an in-person meetup creates a very different kind of connection than reading threads or watching talks later. If you want candid technical conversation, potential collaborators, or simply a sharper picture of where the OSS agent space is heading, this event is a smart place to spend an evening. More broadly, this meetup is useful because open-source AI moves through shared learning. Hearing how others are tackling production constraints can save you weeks of trial and error, while your own perspective may be exactly what someone else in the room needs. Practical Details OSS Agents: Fine-Tuning to Production takes place in person in San Francisco, USA on Wednesday, October 22 at 6:30 PM PDT. The in-person format matters here: this is a meetup designed for real conversation, spontaneous follow-up questions, and networking with people who are actively building. If you are planning to attend, it is worth arriving ready to engage. This is the kind of event where having a concrete problem, recent experiment, or strong opinion can lead to much better conversations. Whether you are working on agent tooling, applications, research, or infrastructure, you will get more out of the evening if you come prepared to compare approaches with others. A few useful things to keep in mind: Location type: in-person meetup in San Francisco Start time: Wednesday, October 22 at 6:30 PM PDT Best for: people interested in AI agents, open-source tooling, production systems, and community connection Format: community-oriented meetup with discussion and networking value built in If the words open source, agents, and production all matter to your work, this event is aimed directly at your interests. Come to learn, pressure-test your thinking, and meet other people who are trying to turn OSS agent ideas into systems that actually hold up in the real world.
Who should attend
This will feel especially useful if you want practical, high-signal conversations about building and shipping open-source AI agents. - You are an **engineer or technical builder** working on agent systems and want to learn how others are handling fine-tuning, orchestration, evaluation, and deployment. - You are a **founder or product-minded operator** exploring how OSS agents can become reliable product features rather than one-off demos. - You are an **ML practitioner or researcher** who wants more exposure to the production side of agent behavior, including tradeoffs that only appear after launch. - You are part of an **AI infrastructure or tooling team** and want to hear what real users and implementers care about when taking agents into production. - You are active in the **open-source AI community** and want to meet other people in San Francisco who are building, testing, and sharing lessons in public. - You are still earlier in your journey but already past the hype phase, and you want grounded conversations that help you understand what it really takes to make agents useful in practice. If you are looking for thoughtful peers, practical lessons, and a clearer view of the path from experimentation to production, you are likely in the right room.