Lessons from AI work.

These are not traditional case studies. They are notes on what becomes visible when AI ideas meet real organizations, real constraints and real work.

CustomGPTs

CustomGPTs for organizational knowledge work

What specialized GPTs teach about knowledge, trust, retrieval and the distance between access and actual use.

Read note

Agents

Agents for real workflows

What agent prototypes reveal about context, handoffs, judgment and the fragile line between support and delegation.

Read note

Skills

Reusable skills for better AI work

What reusable skills teach about turning tacit method into operational memory without freezing the work.

Read note

AI Experiments

Experiments at the edge of useful AI

A chronological field log of CustomGPTs, skills, agents, dashboards and strange tests that reveal where AI becomes useful, fragile or unexpectedly new.

Read log