My AI Stack Is A Workbench
Why I do not treat models as favorites, but as tools with different jobs inside a larger workflow.
Read note ↗AI Workbench
Short technical notes on how AI-assisted work is framed, built, checked, maintained, and sometimes discarded.
Why I do not treat models as favorites, but as tools with different jobs inside a larger workflow.
Read note ↗Prompts help, but reliable AI-assisted work depends on repeatable workflows, standards, examples, and checks.
Read note ↗The useful work starts before generation: framing the real issue, decision, constraints, product, and standard.
Read note ↗I do not use one model for everything. I use a stack, and each tool has a job.
Read note ↗I split review into alignment, verification, assessment, and skeptic passes instead of asking a model to grade itself.
Read note ↗Building the tool is the easy part. Keeping it working is the standing cost.
Read note ↗Not every AI tool earns a place in my workflow. Some cost more than they return.
Read note ↗Good translation is not one prompt. It is a workflow with standards, checks, and human review.
Read note ↗