The Model Context Protocol (MCP) provides a unified method for developers to present tools, data, and capabilities to large language models, establishing itself as a crucial standard in agentic AI. FastMCP, an open-source project managed by the Prefect team, builds on MCP to offer Python developers high-level abstractions to efficiently create and deploy MCP servers and applications. Prefect, an orchestration platform for AI and data workflows, supports this project.
Jeremiah Lowin, the founder and CEO of Prefect, and Adam Azzam, the VP of Product, join Gregor Vand in this episode to explore FastMCP’s origin, its foundational pillars, and the design choices behind FastMCP 3.0.
Gregor Vand, a technologist with a focus on security, formerly served as CTO in various cybersecurity and software engineering firms. Based in Singapore, he can be reached through his profile at vand.hk or LinkedIn.
For the episode’s transcript, click here.
Sponsors:
1. turbopuffer: Companies like Anthropic, Cursor, Notion, Atlassian, and Ramp use turbopuffer for their advanced search features. It’s a cost-effective, serverless vector and full-text search engine. Learn more at turbopuffer.com/sed.
2. Unblocked: Unlocks efficient agentic workflows by synthesizing PRs, docs, Slack, and tickets into an understandable context for agents. Start a free trial at getunblocked.com/sedaily.
3. Retool: Helps engineering leaders manage internal tools securely, providing solutions like AI AppGen for creating secure applications. Explore more at retool.com/se daily.