Published notes
6 ARTICLESLocal LLM Benchmark Method
A repeatable plan for comparing local coding models without invented numbers.
Investigating a PTP Clocking Problem
A structured way to distinguish clock election, network reachability, and audio-flow symptoms.
Safer PowerShell Deployment Checks
Preflight validation, exit-code checks, and post-deployment smoke tests.
A measurement plan for local LLM inference
Separate prompt processing, token generation, memory pressure, and task quality before comparing configurations.
Drawing the boundary between a portfolio and private AI
Why the public website, invited-user chat, and future API have separate access controls and deployment lifecycles.
Troubleshooting across system boundaries
A practical framework for narrowing a fault across clients, networks, applications, and their dependencies.
Room to grow
The notebook will expand through real work across networking, broadcast and media engineering, systems administration, software and automation, AI, and security. New categories appear as useful material is ready.
These initial articles establish the architecture and measurement approach. They do not present private workplace configurations or unmeasured results.