How I Actually Work

Not buzzwords — here's what my day actually looks like.

My day is won or lost in the first hour — so I automated it.
As a Project Manager and Scrum Master running multiple projects, the start of the day — especially Monday — sets the tone for everything after it. Action items, risks, follow-ups, high-priority items, email threads, sprint progress, Slack tags: miss these early and you spend the rest of the day reacting. Done manually, this is hours of scanning before any real work begins. So I built a set of Claude-based automations that hand me the whole picture, insightfully and ready to reference:

  • Daily email digest — a clean summary of overnight communications, surfacing what actually needs my attention from the noise.

  • Slack mention tracker — pulls every tag into a single view of action items, risks, issues, and follow-ups, so nothing slips between channels.

  • Sprint status tracker — fetches tasks, status, progress, and process-compliance in one automated pass — no manual board-scraping.

  • Rolling weekly report — quietly lifts highlights and stats from the above each day, so my weekly status is always written before anyone asks for it.

The point isn't automation for its own sake. It's that I walk into every standup already knowing where the fires are — and I spend my attention on risk and people, not on chasing a status field.

O1. I start the day already informed using AI (Claude)
O2. I run two Agile rhythms at once.

Planned work moves in Scrum — sprint planning, refined backlog, story points, ceremonies. But client and infra reality doesn't wait for a sprint boundary, so ad hoc and support work runs in Kanban alongside it. Knowing which lane a piece of work belongs in is half the job.

O3. I manage and coordinate infrastructure like a product, not a cost centre.

Blue/green deployments, disaster-recovery readiness, monitoring on Grafana, alarms on CloudWatch, error tracking on Sentry, incident alerting on PagerDuty — I give the delivery pipeline the same backlog discipline as a feature. And I watch the AWS bill like it's mine.

O4. I use AI to remove the busywork, not the thinking.

Beyond the morning stack, I've automated process-compliance checks — story points, estimates, due dates, dev checklists — and report generation. I'd rather spend my judgment on what's hard and human than on what a script can catch.

O5. I keep two audiences aligned at once.

Engineering needs clarity and protection from noise. Clients and management need confidence and visibility. A lot of my day is being the translation layer between them — and knowing what each actually needs to hear.