Three Free Lectures That Cover Agentic AI From Basics to Evals

Maven compiled three free video lectures that, taken together, cover the full arc of building with agentic AI: fundamentals, enterprise system design, and evaluation frameworks. All three are available on the Maven platform.

Lecture 1: Intro to Agentic AI

Taught by Mahesh Yadav, a former GenAI Product Lead who has worked at AWS, Meta, Microsoft, and Google. He has 20 years of experience building products across all layers of the AI stack, from chips to LLMs. This lecture covers the fundamental building blocks for building your first agent.

Lecture 2: Designing Agentic AI Systems for Enterprise Use Cases

This is the longest of the three at two hours and 50 minutes. It is taught by a Generative AI lead at AWS and an engineer at OpenAI. The length reflects the complexity of the topic: designing agent systems that work at enterprise scale involves different constraints than building a simple demo.

Lecture 3: How to Set Up Evals for Agents

Hosted by Harrison Chase (CEO and co-founder of LangChain/LangGraph) and Hamel Husain. Evals drive improvement in agent systems over time and ensure everything stays aligned with the intended behavior. This is the piece most builders skip, but it is what separates a demo from a production system.

Key Takeaways

• Start with the fundamentals lecture to understand the building blocks of agents
• The enterprise design lecture (2h 50m) covers the real-world constraints of deploying agents at scale
• Evals are the missing piece for most builders; Harrison Chase's lecture covers how to implement them

Resources

• Maven Lightning Lessons -- Free lecture platform hosting all three courses
• LangGraph -- Framework for building stateful agent workflows, created by the Lecture 3 instructor