Notes on product, applied AI, architecture and engineering leadership.
Lately I open a new app, see they have "added AI" and I already know what I'm going to find: a little bubble in the bottom right corner that opens a chat.
Why Claude Code's output depends less on the model and more on what you write before asking for anything. With two real specs from Venturest.
Nine pregnant women do not make a baby in a month. Development is a system of scarce capacity, and the real leverage lies in what we let in, not in pushing harder on what is already inside.
Real coding use cases, how rules actually work, and what to do with skills and workflows.
And why many are asking the wrong question.
Why 'let's rewrite everything' is almost always the wrong answer — and what to do instead.
A real cost comparison between Twilio Segment and Jitsu, with the numbers that led us to migrate at Venturest and a calculator to estimate your own savings.
After more than 15 years in the tech industry, leading teams in startups and scaleups, I share the technical, human, and personal mistakes that taught me the most.
Si eres buen manager y estás delegando la adopción de IA en tus becarios, lo vas a cagar.
How to move from hype to AI use cases that actually move business metrics.
La semana pasada un amigo me enseñó su nuevo producto. El login estaba hardcodeado en el front. Usuario y contraseña visibles para cualquier...
Tu startup tiene usuarios. Tiene ingresos. Y no sabe por qué crece ni por qué para de crecer.
Trying to innovate with your current team is not a plan. It is a wish.
Rewriting from scratch is almost never the answer. Evolutionary modernisation is.
Fewer vanity metrics, more signals that predict retention and revenue.