AI for Product Management Capstone Exercise — AI for Product Management

What you'll learn

Apply skills from across this course in a single realistic product management scenario, from opportunity evaluation through stakeholder communication
Produce concrete, role-relevant deliverables that demonstrate AI-assisted product judgment across research synthesis, PRD drafting, prioritization, and governance
Self-assess your output against professional quality criteria for each deliverable

Across this course you have worked through where AI creates real leverage in product management and where it does not: synthesizing user research, scanning the competitive landscape, drafting PRDs, evaluating whether an AI feature is worth building at all, prioritizing a roadmap, analyzing feedback and metrics at scale, and communicating AI-informed decisions responsibly. Each of those skills exists to serve one outcome: a PM who can move faster with AI without shipping decisions built on unvalidated shortcuts.

This capstone puts them together in a scenario with the ambiguity, time pressure, and stakeholder dynamics that a real product decision involves.

Capstone Exercise

Evaluating and Scoping an AI-Powered Feature for a Field-Service Scheduling Platform

Context

You are a senior product manager at a 70-person B2B SaaS company that sells scheduling and dispatch software to field-service businesses (HVAC, plumbing, electrical contractors). Your customers are dispatchers who assign incoming service jobs to available technicians throughout the day. In your last quarterly customer advisory call, three separate customers independently mentioned that dispatchers spend a meaningful part of their day manually deciding which technician to assign to each incoming job, weighing technician location, skill match, and current workload. Your CEO, who was on the call, has asked you to 'explore an AI dispatch assistant' and bring a recommendation to the next leadership meeting in ten days. You have access to: a folder of 18 raw customer interview transcripts from the past two quarters (not yet synthesized), your product analytics showing job assignment volume and average time-to-assign per account, and general awareness that two competitors have marketing pages mentioning 'AI-powered scheduling' without much public detail on what the features actually do.

Your Task

Produce four deliverables. First, a one-page research synthesis memo: using the pattern from this course's research synthesis lesson, describe how you would structure an AI-assisted synthesis of the 18 transcripts to validate (or challenge) the dispatch-assignment problem, including what participant-count and quote-attribution standard you would require before treating any theme as real. Second, a feature opportunity evaluation: apply the four-question AI feature opportunity framework from this course to the proposed 'AI dispatch assistant' concept, and state a clear recommendation — proceed as originally framed, proceed with a narrower scope, or do not proceed — with your reasoning tied to specific questions in the framework. Third, if your recommendation is to proceed in some form, a PRD outline: problem statement, three to four goals, a non-goals section you write yourself (not AI-generated), and 4-6 user stories including at least one edge case. Fourth, a one-paragraph leadership briefing: state your recommendation, the evidence behind it, and where AI accelerated your process without overclaiming AI's role, following the communication pattern from the closing lesson of this course.

Your notes (optional)

Deliverable

Four documents: (1) a one-page research synthesis memo describing your AI-assisted synthesis plan and validation standard for the 18 transcripts; (2) a feature opportunity evaluation applying the four-question framework with a clear proceed/narrow-scope/do-not-proceed recommendation and reasoning; (3) a PRD outline with problem statement, goals, a PM-authored non-goals section, and 4-6 user stories including at least one edge case (only required if your recommendation supports proceeding); (4) a one-paragraph leadership briefing stating the recommendation, the sourced evidence behind it, and AI's actual role in reaching it.

Quick check

The first capstone deliverable is a one-page research synthesis memo covering the 18 transcripts. What is that memo supposed to contain?

Select one answer.

Key takeaways

A validated feature opportunity starts with a synthesis standard, not a synthesis output — knowing what would make a theme "confirmed" before you run the analysis prevents an AI-manufactured pattern from driving the decision.
The four-question opportunity framework (validated problem, AI's actual value-add, trust requirement, cost-benefit) applies even under leadership time pressure, and often points toward a narrower, more shippable scope than the original request implied.
A PM-authored non-goals section, grounded in real team and timeline constraints, is what turns a broad concept like "AI dispatch assistant" into a scope a team can actually commit to in a defined timeframe.
A leadership briefing that names its sources and is honest about what remains unvalidated is more credible, and more durable under a skeptical follow-up question, than one that leans on an unearned "the AI confirmed it" framing.