AI Agent System Design for Product Managers
Colin Matthews
Founder, Tech for Product
Learn the foundations of building AI systems
It's surprisingly hard to really learn how AI systems work.
Either you're being told to use a specific framework or tool that you don't use at work, or the materials are way too technical and only an engineer could understand them.
This unique half day workshop takes a different approach than most content you've seen. This is not a course about any specific AI tools or frameworks. Instead, this is a course to help you understand of AI agents one step at a time, starting from the very beginning.
You'll learn how:
1️⃣ The core loop of an agent actually works
2️⃣ Tool calls between the LLM and your product communicate
3️⃣ Improving tool descriptions, errors, and system prompts steer your agent
4️⃣ Giving the agent an environment allows you to build more valuable products
By the end of the session, you'll not only have shipped multiple agents, but you'll also have a new way of understanding how AI product really work.
We'll cover a wide variety of topics related to building AI agents, so please read the full outline below!
Workshop agenda
- 12:00 EDT
How AI agents really work
Learn the basic loop of sending a request to an LLM, giving the LLM some tools to use, and allowing the agent to complete a task
- 12:30 EDT
Improving agents with system prompts, errors, and tool descriptions
Practical steps to improve agent performance
- 01:00 EDT
Monitoring performance with traces and graders
Learn to monitor real usage and manage your agent at scale
- 02:00 EDT
Adding memory and thinking
Determine when it makes sense to add long term memory and thinking to your agent, and how to weigh tradeoffs with cost
- 02:30 EDT
Managing context windows
Get more value out of your context window by compacting prior conversations and pruning out tool call and thinking tokens
- 03:00 EDT
Giving your agent an environment
Give your agent a file system and basic tools to unlock new ways of thinking about AI features
- 03:30 EDT
Retrieving information with agentic RAG
Learn modern approaches to information retrieval with agentic RAG via tool calls
- 04:00 EDT
Special topics
Reserved time for the latest updates in AI!
Learn directly from Colin
Colin Matthews
Founder of Tech for Product, Fractional PM leader
Who this workshop is for
Product managers working on AI products or features
Designers who want to better understand how AI products work
Product leaders who need to lean in to IC work to build their own understanding
What's included
Live sessions
Learn directly from Colin Matthews in a real-time, interactive format.
Lifetime access
Go back to course content and recordings whenever you need to.
Community of peers
Stay accountable and share insights with like-minded professionals.
Certificate of completion
Share your new skills with your employer or on LinkedIn.
Maven Guarantee
Your purchase is backed by the Maven Guarantee.
Frequently asked questions
$650
USD