Skip to content
← All projects
active2024 – Present

AI Management Toolkit

7 custom AI assistants for engineering management — reduced post-meeting documentation from ~45 to ~10 minutes. Built around neurodivergent-friendly, low-cognitive-load workflows.

  • Claude
  • Windsurf
  • ChatGPT
  • Jira
  • Confluence
  • Outlook
AI Management Toolkit
Photo by Planet Volumes on Unsplash

AI changed exactly one thing about how I work: it made automating admin easy. I’ve been automating the boring stuff ever since I got a license — 7 workflows now, one for basically every recurring, boring bit of the job nobody warns you about. Still the best part of AI, if you ask me.

These tools don’t do the thinking. They handle the scaffolding around it — formatting, structure, getting it to the right inbox.

The toolkit

WorkflowWhat it does
/weekly-meetingMeeting notes + stakeholder newsletter from transcript
/capacity-planningQuarterly capacity doc from team data
/rnd-steeringWeekly Jira update for R&D Steering
/goal-companionAnnual goal drafting and refinement
/ambassadorshipTeam ambassadorship competency wiki pages
/meeting-loadWeekly calendar and burnout pattern analysis
/team-docsteam documentation helper

The biggest single win: post-meeting documentation dropped from ~45 minutes to ~10. The AI doesn’t write it for me. It handles the structure, the formatting, and getting it where it needs to go.

Why ADHD-informed design matters here

Every workflow is designed around a principle I’ve learned the hard way: if the tool adds cognitive load, it won’t get used. Each one has a single clear input, a single clear output, and no decisions to make mid-flow.

For a brain that’s already context-switching between two teams and a P1 initiative, that friction removal is the difference between the system working and the system being abandoned after week two.