Episode 26 · Season 1
Top 3 Agentic AI Frameworks (and How to Choose Without Analysis Paralysis)
Agentic AI frameworks are multiplying fast, but most teams only need a clear “when to use what” map to start building. In this episode, we break down the top agentic AI frameworks and how to pick the right one without getting stuck in analysis paralysis.
What you’ll learn
- How LangChain, Pydantic AI, and Anthropic’s Agent SDK differ in structure, community support, and best-fit use cases
- Simple rules of thumb: when you want RAG + ecosystem, typed/structured outputs, or autonomous delegation
- The decision most teams underestimate: why the framework choice is harder to swap than the model
Who it’s for
For product leaders, engineers, and teams building custom agent workflows in production.
Guests
Tim Harting and Gavin Franklin join host Greg Banks to compare their go-to frameworks from real build experience—what works, what’s painful, and what to choose first.
CTA
Follow the show for more practical breakdowns and upcoming agent architecture artifacts.
For questions or feedback, please contact us at CirriusTalk@CirriusSolutions.com
Recommended Links
Link to referenced artifact:
Gavin Franklin on LinkedIn
Tim Harting on LinkedIn
Cirrius Solutions
Cirrius Blog
Apple Podcasts
Spotify
Youtube
iHeartRadio
Greg Banks on LinkedIn
Jason Fowler Music
We cover
- Navigating the Evolving AI Landscape
- Exploring Agent Frameworks: An Overview
- LangChain: The Established Framework
- Pydantic AI: A User-Friendly Approach
- Key Takeaways for Choosing Frameworks
- Anthropic Agent SDK: Cutting-Edge Capabilities
Chapters
- 0:00Navigating the Evolving AI Landscape
- 1:57Exploring Agent Frameworks: An Overview
- 4:54 LangChain: The Established Framework
- 7:43Pydantic AI: A User-Friendly Approach
- 10:14Key Takeaways for Choosing Frameworks
- 15:49Anthropic Agent SDK: Cutting-Edge Capabilities
Resources and downloads
The artifacts Greg’s team built for this episode, free to download.
Top 3 Agentic AI Frameworks (and How to Choose Without Analysis Paralysis)
The episode artifact provides a detailed comparison of three agent frameworks: Anthropic Agent SDK, Pydantic AI, and LangChain. It outlines the key features of each framework and the type of business problem each is best suited to solve.
Related episodes
Agentic AIMCP Servers Deep Dive: Secure, Reusable Tools for AI Agents
Agentic AIBuild Your First Custom AI Agent: A Step-by-Step Walkthrough
Agentic AIContext Engineering: The Secret Sauce Behind Great AI Agents
The Military MinuteJoanah Gacuya – The Volunteering “Hack” That Landed a Military Spouse in SalesforceJoanah Gacuya
Salesforce at ScaleSalesforce at Scale: Tackling Technical Debt Before It Tackles YouTalk to us about AI agents
Keep going: See how we build AI agents for real businesses, Talk to us about a custom AI agent, More from the Cirrius blog