About
One consultant, start to finish.
Placeholder · Your name
Placeholder · Your title
Placeholder · Bio — where you trained and worked
Placeholder · Bio — the data problems you have solved
Placeholder · Bio — why you work this way
Credentials
- Placeholder · Credential or past employer
- Placeholder · Credential or past employer
- Placeholder · Credential or past employer
How I work
01
Definitions before dashboards
Most disputed numbers are disputed because two departments are answering different questions with the same word. That gets settled in writing first. It is the least technical part of the work and usually the part that fixes the most.
02
Loud failures
A pipeline that breaks visibly is a nuisance. A pipeline that quietly serves last month's numbers is a liability. Everything I build asserts on its own freshness and stops rather than publishing something stale.
03
Built to be handed over
Plain Python, tested, in your repository. No proprietary runtime and no per-seat licence that keeps you paying me after the work is finished. The handover is part of the engagement, not an upsell.
04
Accessible by default
Charts are checked for colour-blind separation before they ship, and every one carries a text alternative. This is not a compliance exercise; it is what makes a report survive being read on a phone, pasted into an email, or read aloud.
What I say no to
- Dashboards with no decision attached. If nobody can say what they would do differently, it is decoration.
- Rescuing a number that is wrong at the source. The fix belongs upstream, and I will say so even when the upstream fix is the inconvenient one.
- Work I cannot hand over. If the arrangement requires me to stay indispensable, it is badly designed.
Talk to me directly.
Thirty minutes, no charge. Bring the report you do not trust.