Writing about performance engineering, distributed systems, and the future of AI infrastructure.
December 2024
After building inference infrastructure serving billions of requests daily, I've learned that latency optimization in distributed ML systems requires fundamentally different approaches than traditional web services.
Coming soon
Why every millisecond of latency matters when building AI systems at scale, and how performance constraints shape what's possible in AI product development.
Coming soon
The architectural decisions and optimization techniques that enabled order-of-magnitude improvements in generative AI performance.
Coming soon
The invisible infrastructure layer that makes enterprise AI possible— and why it's more important than the models themselves.