/Events /MCP Community Connect - San Francisco
MCP Community Connect - San Francisco
This event is part of MCP Community Connect, organized by Global AI San Francisco.
Register Now Schedule
About
MCP Community Connect comes to San Francisco for a full day dedicated to building with and for the Model Context Protocol, the open standard connecting AI agents to tools, data, and each other.
Hosted at GitHub's San Francisco office, this gathering brings together the developers, platform teams, and agent builders shaping how MCP gets used in production. Expect hands-on sessions on building and securing MCP servers, talks on client integration and interoperability, and a community showcase of real implementations from the field.
Who's on stage
Tobin works on MCP and AI agents at Anthropic, and is a research fellow at Stanford University, advising research for the Loyal Agents initiative in HAI. Tobin completed his PhD at MIT in 2025 on "Private, Verifiable, and Auditable AI Systems", where he was an Australian-American Fulbright Scholar and a senior fellow with the E14 VC fund. Tobin has been a mentor and advisor across a range of fellowships and institutes, including Pivotal Research, Cambridge ERA, and the Oxford IDAI from the Cosmos Institute. Tobin was an author of the 2025 International AI Safety Report and has research spanning AI security, technical governance, and policy.
Sofía leads partnerships & ecosystems at CopilotKit, where she's driven by a passion for making cutting edge technology more accessible and enjoyable for everyone through rich, beautiful interfaces. Before that, she was a software engineer at Google, working on bringing AI to the company's AR glasses. Originally from Mexico and raised in New York, she's now based in California and spends much of her year traveling — organizing and speaking at events from San Francisco to London, Paris, and Mexico City. Outside of work, she's an avid adventure sports enthusiast, always chasing the next challenge.
Harald Kirschner is a Principal Product Manager at Microsoft working on VS Code and GitHub Copilot AI coding experiences for tens of millions of developers. He is active in the VS Code and developer-experience ecosystem and has spoken on agent memory, MCP, and AI coding tools.
Cedric Vidal is a Principal Cloud Advocate at Microsoft, working on agentic systems and AI evaluation. His focus is on how autonomous AI systems are designed, evaluated, and trusted as they evolve into systems that reason, decide, and act over time.
Prathmesh Patel is the CEO of MCPJam, the evaluation and testing platform for MCP and agent-native software. Previously, he led MCP technical strategy at Asana and led the company’s ChatGPT and Claude app launches in partnership with OpenAI and Anthropic. Led Asana through its public MCP security incident, one of the two biggest enterprise MCP incidents in history. Speaker on MCP evaluation testing at MCP Developers Summit NY 2026.
Jeremy is one of those curious folks with both a technical and a strategic streak. Passionate about people and entrepreneurship, integration and automation. Through technical/business roles at Neo4j, Dagger, GitHub, Twistlock, and Puppet, Jeremy has both zoomed in and zoomed out a lot, acquiring an appreciation for the details and an ever-broader sense of the big architectural picture.
Thierry is a Developer Evangelist at Arcade.dev, focused on the people building production agents on MCP. Previously, he did Data Science at the DHS and Developer Relations at Qdrant. He's spoken at Haystack US, ODSC, and meetups across the AI ecosystem, covering everything from vector search to agentic systems. A former stand-up comedian, data scientist, and teacher, Thierry has made a career of making hard technical ideas land. Born in Burkina Faso, raised in Baltimore, currently based in San Francisco.
Radhika Gulati is a product marketer at Prefect, the team behind FastMCP. She studied computer science at UC Davis and has spent the last three years in data orchestration. Her past lives include hackathon organizer, engineer for a campus food delivery app, and coding instructor for kids.
Liam works on Developer Tools (VS Code / GitHub Copilot) and Model Context Protocol (MCP) at Microsoft and GitHub, helping shape how AI agents talk to the world.
Pamela is a cloud advocate in Python at Microsoft/GitHub. She's been in the tech industry for 15 years, beginning with a role at Google as one of the company's first developer advocates. She went on to become an early full-stack engineer at Coursera and spent many years afterward at Khan Academy, both as an engineer and the creator of computer programming content.
Sessions & tracks
Registration
Main stage ·Welcome
Pamela is a cloud advocate in Python at Microsoft/GitHub. She's been in the tech industry for 15 years, beginning with a role at Google as one of the company's first developer advocates. She went on to become an early full-stack engineer at Coursera and spent many years afterward at Khan Academy, both as an engineer and the creator of computer programming content.
Liam works on Developer Tools (VS Code / GitHub Copilot) and Model Context Protocol (MCP) at Microsoft and GitHub, helping shape how AI agents talk to the world.
Keynote: Evolution of MCP
Harald Kirschner is a Principal Product Manager at Microsoft working on VS Code and GitHub Copilot AI coding experiences for tens of millions of developers. He is active in the VS Code and developer-experience ecosystem and has spoken on agent memory, MCP, and AI coding tools.
Building modern MCP servers with FastMCP
Radhika Gulati is a product marketer at Prefect, the team behind FastMCP. She studied computer science at UC Davis and has spent the last three years in data orchestration. Her past lives include hackathon organizer, engineer for a campus food delivery app, and coding instructor for kids.
Effective MCP: Building reliable stateless MCP Servers
Prathmesh Patel is the CEO of MCPJam, the evaluation and testing platform for MCP and agent-native software. Previously, he led MCP technical strategy at Asana and led the company’s ChatGPT and Claude app launches in partnership with OpenAI and Anthropic. Led Asana through its public MCP security incident, one of the two biggest enterprise MCP incidents in history. Speaker on MCP evaluation testing at MCP Developers Summit NY 2026.
What dozens of enterprise MCP rollouts taught us about security
Jiquan Ngiam is co-founder and CEO of MintMCP, building secure enterprise AI infrastructure that gives AI assistants governed, auditable access to company data, applications, and tools.
Tools, translated: Finding the optimal language for clients & servers
Drew Breunig is the CEO & cofounder of CMPND, editor of O'Reilly's "Context Management Handbook", and writer on technology and culture at dbreunig.com. He thinks AI should be more boring and more useful.
Rendering intelligence: How AG-UI and MCP Apps unlock generative UI for agents
Sofía leads partnerships & ecosystems at CopilotKit, where she's driven by a passion for making cutting edge technology more accessible and enjoyable for everyone through rich, beautiful interfaces. Before that, she was a software engineer at Google, working on bringing AI to the company's AR glasses. Originally from Mexico and raised in New York, she's now based in California and spends much of her year traveling — organizing and speaking at events from San Francisco to London, Paris, and Mexico City. Outside of work, she's an avid adventure sports enthusiast, always chasing the next challenge.
Your product was built for humans. Is it ready for agents?
Cedric Vidal is a Principal Cloud Advocate at Microsoft, working on agentic systems and AI evaluation. His focus is on how autonomous AI systems are designed, evaluated, and trusted as they evolve into systems that reason, decide, and act over time.
Dinner
Main stage ·Your agent has 500 tools. Now what? Using context graphs to route agents through MCP
Jeremy is one of those curious folks with both a technical and a strategic streak. Passionate about people and entrepreneurship, integration and automation. Through technical/business roles at Neo4j, Dagger, GitHub, Twistlock, and Puppet, Jeremy has both zoomed in and zoomed out a lot, acquiring an appreciation for the details and an ever-broader sense of the big architectural picture.
Your tool is in another castle: How five frameworks handle the same MCP server
Thierry is a Developer Evangelist at Arcade.dev, focused on the people building production agents on MCP. Previously, he did Data Science at the DHS and Developer Relations at Qdrant. He's spoken at Haystack US, ODSC, and meetups across the AI ecosystem, covering everything from vector search to agentic systems. A former stand-up comedian, data scientist, and teacher, Thierry has made a career of making hard technical ideas land. Born in Burkina Faso, raised in Baltimore, currently based in San Francisco.
Future of MCP: A fireside chat
After a full day of MCP talks, demos, and discussions, join us for a closing conversation with Tobin South from Anthropic and Harald Kirschner from VS Code. We'll explore what MCP adoption looks like today, lessons emerging from the rapidly growing ecosystem, challenges around interoperability and security, and where the protocol may be headed next. Bring your questions as we reflect on the day's themes and dive into the opportunities and realities of building the connected agent ecosystem.
Tobin works on MCP and AI agents at Anthropic, and is a research fellow at Stanford University, advising research for the Loyal Agents initiative in HAI. Tobin completed his PhD at MIT in 2025 on "Private, Verifiable, and Auditable AI Systems", where he was an Australian-American Fulbright Scholar and a senior fellow with the E14 VC fund. Tobin has been a mentor and advisor across a range of fellowships and institutes, including Pivotal Research, Cambridge ERA, and the Oxford IDAI from the Cosmos Institute. Tobin was an author of the 2025 International AI Safety Report and has research spanning AI security, technical governance, and policy.
Harald Kirschner is a Principal Product Manager at Microsoft working on VS Code and GitHub Copilot AI coding experiences for tens of millions of developers. He is active in the VS Code and developer-experience ecosystem and has spoken on agent memory, MCP, and AI coding tools.
Made possible by
Explore MCP Community Connect
Connecting Agents. Empowering Builders.