BLOG The next AI bottleneck isn't the model. It's the data layer the model has to act through, and that's where the hard, unsolved problems live.
That's the argument behind ** Data Outpost**, a new conference we're bringing to San Francisco on November 4 and 5, 2026. Two days for the people making machine intelligence useful with data inside real companies: founders, engineers, product leaders, researchers, operators, and data architects.
Data Outpost is the next evolution of Small Data SF. Still human, still builder-led, but sharper and more future-facing for the moment when AI starts depending on the data systems underneath it.
We founded MotherDuck on a simple idea: data infrastructure had gotten needlessly complicated, and simplicity wins. That idea became the Small Data Manifesto and the community that showed up to Small Data SF the last two years. That community is now thinking not only about efficiency of their data stacks, but how to adapt their practices and their stacks to take advantage of advances in LLMs. They're also trying to understand how the expertise of data people can enable agentic work. This is the conference to answer those questions.
The program runs on four themes:
Systems: agentic data infrastructure. How will AI depend on the data layer to actually act?
Product: interfaces and workflows. What replaces the dashboard when the interface becomes an operator? How do we build interfaces and workflows for both humans and agents?
Trust: memory, provenance, identity. How do you trust AI-assisted analytics, and build the right context layer underneath it?
Craft: build with taste. What skills will data engineers need, and what role will data analysts play?
It's small-format by design: workshops and talks by and for builders. The Small Data SF stage brought people like Jordan Tigani, Tristan Handy, Barr Moses, and Joe Reis into the room. Data Outpost carries that bar forward, so attendees get real time with the leading builders and innovators in the field.
November 4: Hands-on workshops on leading data + AI tech
November 5: Talks, panels, and plenty of "hallway track"
** Register** to join us for two days of shaping and understanding the future of data and AI.
If you're a builder with a working system or a hard-won argument the market hasn't flattened yet, ** apply to speak**.