Daytona AI Builders - SF, August 2026

An event dedicated to exploring all things AI Engineering! Event partners: Descope & Tensorwave ​Agenda ​​​​​🕒 5:30 pm – 5:35 pmWelcome and Opening Remarks 🎤 Borna Perak, GTM & Partnerships @ Daytona ​​​​​​​🕒 5:35 pm – 5:50 pmTalk "Forkable Sandboxes for Long-Running Agents"​ 🎤 Muhammad Annas Hashmi, DevRel at Daytona ​​​​​​​Outline: Agents are starting to do real software work: installing dependencies, editing files, running tests, starting servers, and debugging over long sessions. But most AI systems still treat each run like a disposable prompt-response loop. That breaks down once the useful state lives not just in tokens, but in the actual machine.This talk is about the runtime layer agents need to become reliable builders. I’ll show how forkable Daytona sandboxes let you pause at important moments in an agent run, branch the environment, and sample multiple continuations from the same machine state. For builders, this turns long-running agent workflows into something you can inspect, compare, recover, and eventually optimize. ​🕒 5:50 pm – 6:05 pmTalk "Identity Tips to Secure AI Agents and MCP Servers" 🎤 Mrunank Pawar, DevRel Engineer at Descope ​​​​​​​Outline: Every organization has an agentic AI / MCP project, but how many of these are in production? And if they aren't in production, how likely is it that authentication, access control, and agentic identity concerns are the reason? This session will share real-world MCP use cases, common agentic identity challenges, and practical identity tips to secure AI agents and MCP servers. ​🕒 6:05 pm – 6:20 pmTalk "Building a real time semantic search engine from agent transcripts" 🎤 Evis Drenova, Principal Software Engineer at Entire ​​​​​​​Outline: Research has shown that giving agents context on previous work done improves their performance. But how do you find the context that agents need in a reliable and efficient way when you have millions of agent conversations? This talk goes through how we built a reliable and scalable real time semantic search engine that serves relevant context to agents (and humans) from millions of agentic coding sessions. ​🕒 6:20 pm – 6:30 pmTalk "Building Programmable Review Agents for Agent-Written Code" 🎤 Zhuolun Li, AI Engineer at Macroscope ​​​​​​​Outline: Coding agents are making code generation dramatically cheaper, shifting the bottleneck from writing code to verifying whether it's safe to ship. We'll live-demo a PR reviewed by two review agents: a general-purpose Code Review Agent that catches correctness issues, and a programmable Check Run Agent that enforces specific rules. We'll then share what we learned building review agents that run automatically on every PR—especially around balancing recall and precision, investigation depth, and deciding what's actually worth reporting. ​🕒 6:30 pm – 6:40 pmTalk "Build Fresh data view for long horizon agents" 🎤 Linghua Jin, Co-founder and CEO of CocoIndex ​​​​​​​Outline: Agents run ~50x faster than humans and create massive unstructured data. You cannot fit the entire thing into the agent; it requires building a view over a large unstructured corpus. CocoIndex is a context engine for agents: a new computing paradigm over evolving unstructured assets and multi-stage processing. You declare the state of your target as a function of the source; the engine keeps it in sync with minimal reprocessing, across time horizons, at scale. We will go over the topics around how to build fresh data view for long horizon agents and do live demo around it. ​​​​​​​​​​​​​​​🕒 6:40 pm - 8:00 pm ​​Networking With pizzas and beverages About event This is dynamic gathering for AI enthusiasts, innovators, and professionals to collaborate, share ideas, and explore the latest advancements in artificial intelligence. Whether you're building AI products, researching cutting-edge algorithms, or simply passionate about the field, join us to connect, learn, and drive the future of AI…