Tyler Collins
Research computing consultant, Canada
I help researchers turn computational problems into reliable software and workflows: HPC systems, Python tooling, EEG/MEG pipelines, getting more out of the hardware you already have, and helping scientific teams use AI coding agents well.
Recent writing
All writing-
Writing an AGENTS.md for a national cluster
What I learned putting an AGENTS.md into the Alliance software stack: skip what frontier models already know and spend the lines on site policy and when to stop and ask.
-
Your LLM Loop Is a Batch Job
800,000 small requests in a day slows a shared LLM service down for everyone. If your workload looks like that, a batch job on a GPU node will finish sooner.
-
Start your five-hour window early
How Claude and Codex five-hour and weekly limits reset, and why an early-morning ping can get you a fresh window by lunch.
-
Tectonic makes LaTeX easier to work on with an agent
Using Tectonic and a coding agent to edit LaTeX, compile a résumé, and check the rendered PDF without browser automation.
-
When to reach for Opus
Sonnet works well when I break down the work for it. For a whole feature ticket with decisions along the way, I'd try Opus.