Training
LLMs in product, for PM teams
A one-day workshop for product managers to understand GenAI software development lifecycle. Focused on the parts of the work that look different — prototyping, evals, rollout — not on prompting.
- Audience
- Product managers and product leads
- Format
- One-day workshop, on-site or remote
- Duration
- 1 day
Who this is for
Product managers who are now responsible for AI features or work in teams that follow agentic coding to ship new features. We worked with companies like Sony Interactive and others who realized that the playbook from the deterministic-software world doesn’t quite apply. You don’t scope a feature but prototype it yourself; you adapt test cases to become evals; the fundamentals of customer understanding and knowing what to build don’t go away but we help you reframe them.
What we cover
- Scoping AI features: how to define “done” when the success criterion isn’t a passing unit test.
- Evaluation in practice: what an eval suite actually looks like, who writes the cases, and how to keep it from rotting.
- Rollout patterns: shadow mode, gated rollouts, and the cost of getting it wrong for AI features specifically.
- Cost and latency as product variables: when they matter, when they don’t, and how to talk about them with engineering.
- The “is this a feature or a product” question: when an LLM capability should ship as a feature inside an existing product vs. as its own thing.
What we don’t cover
- Prompting tricks. There are good resources for that already and we won’t waste a day on it.
- Generic AI strategy. We assume your team has already decided to ship AI features or adopt agentic coding practices; we’re focused on how to ship them well.
Format
One day, customized to the product context. We ask for a short overview of the features in flight, and we use those as the working material throughout the day. Limited to 8 people.