The Model Context Protocol (MCP) offers developers a standardized approach to make tools, data, and capabilities accessible to large language models, rapidly becoming a key standard in agentic AI. FastMCP is an open-source project managed by the team at Prefect, an AI and data workflow orchestration platform. This project expands upon MCP, delivering high-level, user-friendly abstractions for Python developers to swiftly build and deploy MCP servers and applications.
Jeremiah Lowin is the Prefect founder and CEO, while Adam Azzam serves as the VP of Product. In this episode, they join Gregor Vand to explore the FastMCP origins, its three core pillars, the design choices in FastMCP 3.0, and more.
Gregor Vand, a security-focused technologist and former CTO in various tech companies, is based in Singapore and reachable via his profile at vand.hk or on LinkedIn.
Check out the episode transcript here.
Sponsors
Turbopuffer helps companies like Anthropic and Atlassian efficiently handle ambitious search features with a serverless vector and full-text search engine, offering significant cost advantages and speed. Visit turbopuffer.com/sed for a free month.
Unblocked aids coding agents in gaining context, improving decision-making, and code quality by synthesizing organizational knowledge. Try it free for three weeks at getunblocked.com/sedaily.
Retool provides a controlled environment for building tools, preventing shadow IT and maintaining security. Discover more at retool.com/se daily.