Insights for the Rest of Us
Analytics and AI without the intimidation. Plain-language guides to SQL, data work, and practical AI — examples you can use at work, a little humor, no PhD required.
Latest posts
All articles →
Multi-step jobs and checking results
Part 3 of the ChatGPT Work tutorial. Run multi-step jobs as a goal-plan-act-checkpoint loop: review the plan, steer mid-run, check usage, and finish with a result checklist that catches workslop before anyone else sees the file.
Markdown and project instruction files (CLAUDE.md, AGENTS.md, and friends)
Part 6 of the Claude Code tutorial. How CLAUDE.md, AGENTS.md, README, and related instruction files split work between humans and agents, with…
What bad data actually means
Bad data is a shrug, not a diagnosis. Learn five practical quality dimensions (completeness, accuracy, consistency, timeliness, uniqueness) with workplace examples so…
First agentic task with approvals
Part 2 walks one low-risk agentic task with approvals: owned notes into a short status deck, tight tool access, and human review…
Memory and context: what sticks, what resets
Part 5 of the Claude Code tutorial. Chat resets with /clear or a new session, CLAUDE.md reloads every time, auto memory can…



