AI productivity practice exercise for product managers

What is the challenge scenario?

Marcus Osei is a product manager at Fielder, a 200 person B2B SaaS company that builds project management software for construction firms. He has been in the role two years and he is good at it. He is also stretched thin.

His head of product, Priya, just came back from a conference where every session pitched AI as a 10x productivity lever. She told Marcus to audit his own workflow, figure out where AI genuinely helps, and come back in 48 hours with something concrete, not a slide deck.

Marcus pulled his time log from the last two weeks. Six recurring tasks show up: a weekly stakeholder update email that takes 2.5 hours, user interview synthesis at 4 hours, a sprint planning brief at 3 hours, competitor monitoring at 2 hours, a feature prioritization session at 3 hours, and incident post mortems at roughly 1.5 hours each, twice a month.

He has heard of the Jagged Frontier, the idea from a BCG and Harvard study that AI is surprisingly strong at some tasks and surprisingly weak at others, with the pattern rarely obvious in advance. He knows that pointing a chatbot at everything is not a plan. Priya has no patience left for abstract frameworks.

What are you asked to do?

Step into Marcus's position and produce four things inside the 48 hour window. First, classify all six of his tasks against the Jagged Frontier using four questions: how much the task depends on pattern recognition, creative recombination, judgment of genuinely new situations, and emotional intelligence. Second, pick the three tasks where AI has the most impact potential and decide, for each, whether Marcus should work Centaur style with clearly divided roles or Cyborg style with a tight, embedded loop, naming the strongest counterargument to your own choice. Third, design a five stage adoption plan for the single highest value task, specific enough that Marcus could start stage two tomorrow morning. Fourth, write the actual prompt Marcus would paste into Claude or ChatGPT, including placeholders for real content. Close with a 150 word reply to Priya that names the task, the time saved, and how Marcus will check in a week that it worked.

What does a strong answer cover?

Since this challenge has no published rubric, judge your own answer against the questions a strong submission answers along the way, drawn from the bootcamp's own curriculum:

• Does the answer design an AI workflow around one of Marcus's real tasks, not a generic productivity tip?
• Does it separate what AI should draft or automate from what still needs Marcus's own judgment or relationship work?
• Does it name a concrete way to turn a meeting or interview into action items without manual cleanup afterward?
• Does the research and synthesis step end in a decision, not just a longer summary?
• Does the plan include a real check, a week out, for whether the change actually worked?

Skills this exercise practises

Task triage under a real framework, choosing between two distinct human plus AI working styles, designing an adoption plan with a built in validation step, and writing a complete, structured AI prompt rather than a one line request.

Which bootcamp does this challenge come from?

This is the practical challenge from 10x Productivity with AI, Builders Camp's one week bootcamp on building a personal AI system that compounds, led by Andre Albuquerque across two 90 minute live sessions. Completing it counts toward that bootcamp's certificate, the same certificate every completed Builders Camp bootcamp issues. If you want more structured practice with AI as a working partner rather than a novelty, the best AI tools for product managers guide and the ChatGPT for product managers field guide cover the same ground from a tool selection angle, and how to become an AI product manager puts this specific skill into a longer career context.

Builders Camp runs live and self paced bootcamps in product management and AI product building. See the 10x Productivity with AI bootcamp for the current cohort dates and the full self paced curriculum.