AI product discovery requires a unique approach, blending problem-solving with deep technical, data, and ethical considerations. Learn how to structure sprints that deliver actionable insights and viable AI solutions.
Read article →
Product Craft
Audio
AI product discovery requires a unique approach, blending problem-solving with deep technical, data, and ethical considerations. Learn how to structure sprints that deliver actionable insights and viable AI solutions.
AI Product
Audio
As AI Product Managers, understanding how to best customize large language models is critical. This guide breaks down RAG and fine-tuning, helping you choose the right approach for your product's specific needs and constraints.
Metrics
Audio
Selecting the right evaluation metrics for your AI feature is one of the most critical decisions a Product Manager makes. It dictates success, guides development, and ensures you're building real user value.
AI Product
Audio
Writing a Product Requirements Document (PRD) for an LLM-powered feature demands a unique approach. This guide helps AI Product Managers navigate the complexities of probabilistic systems, data strategy, and ethical considerations.
AI Product
Audio
Trust is a product surface, not a compliance checkbox. How I design confidence signals, graceful fallbacks, and human-in-the-loop guardrails so users lean on AI instead of second-guessing it.
Personalization
Audio
A single model rarely moves the needle alone. The real lift comes from treating personalization as a system - signals, ranking, feedback loops, and the experiences that make them legible to the user.
Product Craft
Audio
The spec is no longer a static document - it is an evaluation harness. How I frame outcomes, edge cases, and eval criteria so an AI feature ships with a way to prove it works.
Agents
Audio
High containment is not about resolving everything - it is about knowing precisely when not to. Lessons from building conversational agents that escalate cleanly and keep trust high.
Metrics
Audio
Accuracy is a vanity metric if no one changes behavior. The metrics I actually watch - task success, deflection, time saved, and the leading indicators that predict retention.
AI Product
Audio
Prompts are product decisions. A practical framework for PMs to reason about instructions, context windows, and failure modes without needing to write the model code themselves.