Writing
Notes on product, AI, data, growth, technology, and lessons from building.
Starting with the problem
A note on why problem definition often matters more than the feature that follows.
What makes an AI product genuinely useful?
Thoughts on designing AI around real workflow friction rather than simply adding a conversational interface.
Why governed metrics matter for AI
Why semantic layers, metric definitions, and trust become more important as analytics interfaces become more natural.
When should an internal tool become a platform?
How to recognize reusable capabilities without turning every product into generic infrastructure.
What analytics taught me about product management
Lessons from moving from analyzing decisions to being responsible for making them.
Building the stack behind AI
Thinking about the interaction between hardware, infrastructure, models, software, and applications as AI scales.