Discover and Define High-Impact AI Agents with a Practical Framework

Amir Piltan

AI Product Leader

AI is moving fast. Knowing what to build is still the most important first step.

AI is evolving fast, opening up new possibilities rapidly. But for product and business teams, the hardest question often isn’t “How do we build it?”, it’s “What should we build in the first place?”

You may have a business problem or workflow where AI could help, but still struggle with:

  • What would the AI Agent-powered workflow look like?

  • Which parts should AI automate vs. augment?

  • What data and tools would it need?

  • What could go wrong?

  • Is an AI Agent the right solution?

  • Which opportunities should we prioritize?

  • How can we quickly test feasibility and validate?

There are plenty of resources on how to build AI Agents. This course focuses on the product work that comes before: discovering, defining, evaluating, prioritizing and validating right opportunities before investing heavily in development.

You’ll learn a practical and structured framework you can apply to a real-world workflow, moving from problem to opportunity to a well-defined AI Agent concept to validation, each step building on the last, with examples, templates, and AI Copilot prompts along the way.

Grounded in a decade of building AI products at scale, the framework combines product & design thinking with AI tech understanding.

What you’ll learn

Turn real manual workflows and user pains into well-defined, evaluated and prioritized AI Agent concepts ready for validation or a PoC.

  • Start with business and customer problems, rather than AI technology.

  • Map the current user journey, including Jobs To Be Done, user actions, system interactions, pain points and bottlenecks.

  • Identify opportunity areas and turn them into focused How Might We questions.

  • Define what the user and agent do at each step of the workflow.

  • Identify the data, tools, systems and outputs the agent needs.

  • Decide where AI should automate, augment or leave the user's work unchanged.

  • Identify potential errors and failure modes across the AI Agent journey.

  • Cluster and assess failures as tolerable, risky or unacceptable.

  • Use failure analysis to uncover risks, gaps and areas requiring further validation.

  • Determine the desired interaction between the user, agent, tools and systems.

  • Explore relevant agent patterns such as tool use, reflection and other approaches.

  • Match the agent pattern to the use case without introducing unnecessary complexity.

  • Run a Reality Check across required data, tools, systems and technical feasibility.

  • Evaluate AI fit, user-flow synergy, business value and risk.

  • Prioritize opportunities and identify critical assumptions to test through prompting or a lightweight PoC.

  • Define meaningful user and business success metrics.

  • Establish quality evaluation metrics, approaches, thresholds and frequency.

  • Create an evaluation framework to guide prototyping, development and iteration.

Learn directly from Amir

Amir Piltan

Amir Piltan

An AI Product Leader with 10+ years of experience in AI Product delivery

See all products from Amir Piltan

Who this course is for

  • AI PMs, domain PMs and designers who are exploring where AI Agents can create meaningful value for their product.

  • Domain experts and business leaders who know their problems and workflows and want to turn them into AI Agent opportunities.

  • AI transformation professionals working with teams to discover and define AI-powered solutions.

What's included

Lifetime access

Revisit course content, recordings, templates and resources whenever you need.

Framework + practical templates

Get the complete AI Agent framework and ready-to-use templates to apply it.

AI Agent Framework Copilot

Use custom AI prompts to work through the framework and challenge your thinking.

Homework Assignments

Put the framework into practice with optional focused assignments between sessions.

Additional Office Hour Q&As

Get extra time for questions, challenges and personalized guidance.

Peer discussion & feedback

Share ideas, compare approaches and learn alongside other students.

Certificate of completion

Share your new AI Agent product skills with your employer or on LinkedIn.

Maven Guarantee

Your purchase is backed by the Maven Guarantee.

Course syllabus

Week 1

Nov 3—Nov 8

    Foundation & Opportunity Mapping

    2 items

Week 2

Nov 9—Nov 15

    Concept Development (part 1)

    2 items

Frequently asked questions

Maven for Teams

Reimbursement

Get your company to pay

Everything L&D needs: email template, receipts, and certificate of completion.

Get reimbursed

Private cohort

Run a cohort for your org

A dedicated cohort with a custom schedule and curriculum, tailored to your team.

Book a private cohort