Insights
Notes from building across disciplines.
First-hand lessons from inventing products, building production AI, leading companies, and working across software, studios, and stages.
Brand is part of the product
Why positioning, interface design, system behavior, operations, and perceived value have to work as one coherent product.
Designing an AI fraud-prevention system people can trust
A practitioner’s framework for turning uncertain signals into proportionate decisions—and every reviewed outcome into a better system.
Designing an AI product for the next step
Clear decisions, precise states, source-connected explanations, explicit uncertainty, and human control in high-stakes AI products.
Building a social product before social media
What the surviving Tegnebordet archive reveals about product loops, design, moderation, trust, and operating an online creator community as a teenager.
AI that survives contact with production
Decision design, evaluation, human control, security, and the operational work that turns a capable model into a dependable product.
From unmet need to working product
How workarounds become evidence, prototypes become systems, and direct customer exposure creates better product judgment.
Building tools for experts
How to preserve capability while making complex software feel clear, predictable, reversible, and trustworthy.
What music production taught me about software teams
Taste, signal flow, shared language, decisive iteration, and why removing work can be the highest-leverage move.
Constraints create product discipline
What real operating constraints teach a team about focus, automation, ownership, and sustainable execution.