Entry #005
Distributed State in Large-Scale LLM Orchestration
How we held sub-50ms consistency across regions for ephemeral model state — and why most teams reach for consensus protocols they do not need.
AI 12 May 2026
The Archive
A clinical examination of applied AI, software architecture, neurodiversity and human performance — every entry, in one place.
5 entries shown Sort: recent
Entry #004
Why Your LLM Eval Harness Is Quietly Lying to You
Offline eval scores that climb while production quality flatlines are the default failure mode of applied AI. Here is how the gap opens, and how to close it.
AI 29 Apr 2026
Entry #003
The Hidden Cost of Microfrontends in Finance
Microfrontends promise team autonomy. In a regulated finance product they quietly traded one shared codebase for a distributed governance problem nobody owned.
Engineering 18 Mar 2026
Entry #002
The Hyper-Systemizer — Engineering Daily Life on Protocols
Raising an autistic child taught me to treat the home like a system with interfaces and failure modes. Here is what that reframe changed, honestly.
Autism 26 Feb 2026
Entry #001
Training Physical Capacity for Deep Engineering Work
After fifteen years at a desk, I started treating training like infrastructure for cognition. The protocol is boring, measurable, and it works.
Fitness 21 Jan 2026
No entries in this category.