Make Something Agents Want
Sherveen Mashayekhi
Founder, Free Agency & AI Muscle
A new customer is replacing humans: AI agents.
Hey there. We're at the beginning of the Agent Experience (AX) layer -- products like Claude Code, OpenClaw, Codex, + Manus are showing us early versions of a new business paradigm:
Moving forward, your customer is both:
your principal: a human at the other end with a goal
your operator: an AI agent operating on their behalf
More and more, AI agents will be recommending products to their users, navigating between options, executing a transaction, and even using your service or product an ongoing basis.
And you HAVE to build differently for agents.
Unlike humans, they don't buy from emotion-driven landing pages. They don't have eyes that prefer bright colors to draw their attention to next steps. They definitely prefer to avoid your interface if they can avoid it.
Companies like Stripe and Salesforce are going agent-first with their releases, startups like AgentMail are building toolkits to help agents become more autonomous and capable, and OpenAI & Anthropic are embracing long-running agents as the killer use case of AI models.
Learn how to build your product or service around the agent lane for AI, the new human lane that keeps people in mind, and the control lane that makes it all work.
What you’ll learn
Learn about the tools, behaviors, systems, & business models powering the AI agent future, and how to steer your products + services to win.
The technical protocols that fuel the most proactive AI agents, incl. the differences between competing frameworks and visions of the future
How agents think, see, and act: APIs, CLIs, MCPs, browser + computer use, and other ways in which agents view and transact toward their goal
The human operator: what's different or the same for the human behind the agent, and how do you serve both stakeholders at the same time
How to build an agent-first and LLM-first layer to your human UI, making it easy and feasible to continue to serve both
Differences in product architecture, including observability, product instrumentation, evals, idempotency, and auth
Agent frameworks driving the ecosystem (ex. Mastra, LangChain, Claude SDK) and how to ensure your agent lane serves all agent types
Learn about the emerging frameworks helping agents pay and move money around: Agent Commerce Protocol (ACP), Machine Payments Protocol (MPP)
Agents don't want to use your product the same way a human does -- how do you expose the right features, the right way?
public web → structured extraction → browser automation → native tools / APIs / MCP → inter-agent / commerce protocols
The shift from business-to-human to business-to-agent, and the model for product adoption, retention, and transaction
Metric inversion: time-on-page no longer matters, but agent guardrails and API menu dynamics do
How your business's value proposition changes in an AI agent world, and how to price for it
What will change when Agent-to-Agent (A2A) protocols gain popularity (est. late 2026 to 2027)?
Why did OpenClaw spike so hard, but fall so far? Lessons from the first proactive agent
Long-running agents, asynchronous agents, and fully autonomous agents
How to cater for agents across every surface area: marketing, operations, sales, product, engineering, + more
Implementation and design of the agent lane, including read/do/prove, identity, auth, traces, guardrails, and privacy
What to try: live the future to understand it -- the products and businesses already serving an agent lane,
Learn directly from Sherveen
Sherveen Mashayekhi
Founder @ Free Agency + AI Muscle. AI product leader, angel investor, supernerd, and coach.
Who this course is for
Product leaders, designers, marketers, and engineers looking to ensure their team doesn't get beaten by an agent-first competitor.
Founders and CEOs of startups, software companies, and services businesses who want to serve agents as they become the #1 customer.
Builders who are building autonomous AI agents looking to ensure their capability and compatibility in the ecosystem.
Prerequisites
No technical background required, though technical concepts are covered!
You'll walk away: understanding how agents behave (marketers), adopt your product/business (product), or implement with nuance (engineers).
What's included
Live sessions
Learn directly from Sherveen Mashayekhi 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.
Async resources + projects
Lifetime access to guides, reading, recommendations, and other resources.
Maven Guarantee
Your purchase is backed by the Maven Guarantee.
Course syllabus
Week 1
Week 2
Jun
8
Transactions: Identity, Permissions, & Payment
Jun
9
Agent Economics: Business Models & Integration
Jun
10
Agentic Futures: A2A, Emergence, & Evals
Technical Implementation: How to Build + Test for Agents
Trust UX: Keeping the Human at the Center
Free resource
How to Build Your Product for AI Agents: 2026's New Customer
AI agents are browsing, clicking, and converting
2026's new customer isn't always a person. What do AI agents in the wild mean for your product or business?
How to make your product agent-friendly (without a rewrite)
Best practices for ensuring that AI agents like your product, recommend it, and use it well.
The technical spectrum: an overview
APIs → structured extraction → browser automation → agentic browsers. What breaks where, and why.
Schedule
Live sessions
3-5 hrs / week
All sessions are recorded and projects available to do on your own schedule (lifetime access).
Mon, Jun 1
6:00 PM—7:00 PM (UTC)
Tue, Jun 2
6:00 PM—7:00 PM (UTC)
Wed, Jun 3
6:00 PM—7:00 PM (UTC)
Projects
1-6 hrs / week
Optional -- do projects at your own pace, on your own time.
Async content
1-6 hrs / week
Optional -- go through async material at your own pace, on your own time.
Frequently asked questions
$1,500
USD