A beginning

From search, ads and operating systems to large models and agents, the shape of technology keeps changing. The essence of engineering does not: building dependable systems within real constraints.

Many of the judgments and lessons worth keeping have lived only in internal docs and meeting rooms. This is where I write them down.

What you will find here

• AI infrastructure: training stability, inference scheduling, rate limiting and quotas — the engineering that decides how far intelligence can go.
• Agents: the architecture of desktop agents, the execution loop, and the trade-offs behind skill frameworks.
• AI-native organizations: how large models and coding agents reshape R&D, design, operations and supply chain.
• Engineering leadership: building teams from zero, developer productivity, and working across regions.

I am always glad to hear from readers — write to me.