You're not building for humans anymore. You're building for their agents.
Your customers are starting to use agents and they’re asking if your product can be used by them. They'll eventually stop using your UI directly.
If you don’t have a plan, you won't be ready when agents start deciding which products get used, and which get skipped.
You already recognize this, but you need the know-how to plan your roadmap. Will you rely on unreliable web scraping, or give agents what they really want: MCP servers, easy-to-use APIs, plugins in ChatGPT and Claude, and Skills? Which of those your product needs, and in what order, is a product decision.
If you’re nodding yes, this course is for you: 5 sessions where we learn, build, and prototype every surface agents work with and show which are best for your use case.
After the course, you’ll meet 1:1 with the instructor for feedback and questions. A month later, we'll meet again as a group to show what you built and peer review.
You don't need to be a coder, we'll use AI for that, and you'll get credits for Claude Pro, Replit, and Fractal - nothing extra to buy.
You’ll be the person who's built and tested products with agents, seen where and why they fail. And you'll have the tools and frameworks to build what agents want.
What you’ll learn
Become a hands-on product builder of what agents can use: MCP, APIs, Plugins, Connectors, and Skills.
Map how APIs, MCP, CLI, Plugins and Skills enable agents.
Select the right mix for your product and use cases.
Understand why you should avoid depending on agents scraping your UI.
Test what works and what doesn't work for agents - in Claude, live.
Watch a live agent skip the docs, guess from training data, and recover from good error design.
Leave with an agent-readiness checklist and an API design Skill - to audit the APIs you have or design the ones you don't.
Generate an MCP server from an existing API, and then build one from scratch with Replit.
Write the MCP metadata that acts as agent UX, which drives LLM usability - tool names, descriptions, and annotations.
Measure task success rates against prompts to build a baseline so you can establish your own evals.
Experience how chat apps render specialized UI right into the conversation with users.
Build a shippable MCP app including its interactive UI using no-code AI tooling.
Get ready to submit, and know what OpenAI and Anthropic require for submission to their directories.
Dissect a Skill's anatomy - the SKILL.md file, progressive disclosure, triggers, and the five product use cases.
Write, security-scan, and test a Skill from scratch for your own product during the session.
Submit your Skill to a directory for customer distribution and understand what drives installs.
Decide which mix of APIs, MCP, CLI, Plugins and Skills is appropriate for your customers.
Assemble your artifacts as Connectors and Plugins and prepare for distribution.
Meet 1:1 with your instructor once the course is over - and then once more with the group to show what you built.
Who this course is for
Product Managers taking their product to the next level - so it works when customers start using agents instead of clicking the UI.
Platform and API PMs deciding which of their APIs to optimize or turn into MCP servers, Plugins and Connectors - and in what order.
Founders & Executives who want to transform their companies and fully leverage AI opportunities and a market of +1.3B users.
Prerequisites
Some API knowledge is required
This is a fast-paced, hands-on course, and while you don't need to be a coder, it helps to understand HTTP basics or to have shipped an API.
If you need to brush up on your API skills:
Consider taking API Fundamentals for Product Managers first: https://maven.com/emmanuel/api-pm-fundamentals
Some AI experience is expected
We'll walk you through building with LLMs and AI tools, even if you've never used them before. Familiarity with Claude/ChatGPT is a plus.
What's included
Live sessions
Learn directly from Emmanuel Paraskakis 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.
$109 of credits for Claude Pro, Replit, and Fractal so you can build
We'll include a month's worth of credits plus access to free tools, so you won't have to buy anything extra.
A set of design and evaluation Skills for APIs, MCP, CLIs will be included
Install the Skills and use them with your requirements and standards docs to design, build, and evaluate agent-friendly products.
Agent-ready API checklists and API design intro docs are part of the course
Review the checklists and customize them for your own organization. Use them as a starting point for documents you can share with your agents and team.
A one-hour deep dive with the instructor for feedback and deep dives
Get personalized advice on topics such as career, product strategy and roadmaps - or where to take your learning path next.
A group demo session and retrospective a month after the cohort ends
Show your peers what you built, share your learnings, and get your questions answered after you've had a chance to digest and use your knowledge.
Maven Guarantee
Your purchase is backed by the Maven Guarantee.
Course syllabus
5 live sessions • 29 lessons
Week 1
Sep
8
Session 1
Tue 9/84:00 PM—7:00 PM (UTC)Sep
10
Session 2
Thu 9/104:00 PM—7:00 PM (UTC)
Module 1 - Build What Agents Want
Pre-Reading Materials - Module 2
Module 2 - MCP Deep Dive
Free resource
Master MCP Essentials for PMs
Explain MCP in plain PM language
Understand clients/servers, tools/resources, and how MCP lets agents and LLMs "see" your data.
Spot product bets MCP unlocks
Map MCP to agents and connected apps, API strategy, and what it enables for ChatGPT experiences and Store-ready apps.
Assess spec maturity and risks
Know what’s stable vs evolving, plus auth, security, and how governance is shaping the protocol.
Try MCP inside an LLM (hands-on)
Add a remote MCP server, and run a few tool calls end-to-end, so you can brainstorm product ideas for the future.
Where to find the best next resources
Leave with the key docs, links, resources, and communities to follow as MCP evolves in 2026.
Schedule
Live sessions
9-10 hrs
Tue, Sep 8
4:00 PM—7:00 PM (UTC)
Thu, Sep 10
4:00 PM—7:00 PM (UTC)
Tue, Sep 15
4:00 PM—7:00 PM (UTC)
Frequently asked questions
Maven for Teams
Reimbursement
Get your company to pay
Everything L&D needs: email template, receipts, and certificate of completion.
Get reimbursedTeam discount
Learn with your teammates
Save 20%+ when 2 or more teammates enroll in the same cohort.
Save 20%+ with a teamPrivate cohort
Run a cohort for your org
A dedicated cohort with a custom schedule and curriculum, tailored to your team.
Book a private cohort$899
USD