Skip to main content
/
Embedded transformation / software delivery

Inside the work.
Sharper delivery.
AI with judgment.

I work embedded with the people building the software and stay close to the people accountable for the outcome. My focus is the gap where leadership intent, team reality, process, and AI adoption have to become one workable delivery system.

thesis /

from intent to team-level execution

Transformation usually fails in the middle: goals get softened, incentives pull apart, and teams receive contradictions too late. AI adoption has the same problem. The durable value will come from clear work design, trusted feedback loops, useful assistants, and deterministic tooling that teams can actually own.

Let's talkSelected workRead articles

20+

Years of Experience

19

Client Engagements

16

Companies

2008

Projects Since

Position

this is
  • +working embedded in teams that build the product, not advising from the outside
  • +connecting leadership intent with code, reviews, planning, incentives, and trade-offs
  • +treating agile, process, and AI as connected transformation work
  • +direct, honest consulting with respect for leaders, middle layers, and the people doing the work
this is not
  • -AI influencer positioning
  • -cargo cult agile or AI adoption
  • -premature standardization from above
  • -slide decks detached from delivery

Operating model

practice

Keep intent connected

I work in the team while keeping the leadership intent visible. The job is to stop good goals from being watered down before they reach planning, code, review, and release habits.

Signals: goal clarity, decision latency, trust, WIP, handoffs
practice

Align the operating system

Process change only sticks when incentives, ownership, feedback loops, and architecture point in the same direction. I help make the real system visible enough to change.

Signals: ownership gaps, escalation paths, review depth, rollback rate
practice

Make AI adoption usable

AI is useful when it becomes part of the work system: clear use cases, local or private assistants where they make sense, deterministic tooling around them, and human review where judgment matters.

Signals: context quality, review load, cost, risk, repeatability
From the workbenchrecent notes
nowshapingwhy agile and AI transformations run into similar tensions
this weekreviewinghow trust and sponsorship change team-level adoption
soonwritingwhat premature standardization breaks before it helps
alwaystestingclean code as the base layer for useful AI

Topics

active

Clean code and safer change

Boundaries, naming, tests, refactoring, and code that stays understandable under pressure.

notes and essaysread
active

Practical agile inside teams

Learning speed, trust, alignment, slicing, WIP, and the habits that make adaptation possible without relying on ceremonies alone.

operating notesread
evolving

Delivery with AI in the loop

AI adoption as embedded transformation work: leadership intent, team habits, clear work packages, local assistants, deterministic tooling, review points, and ownership that grows from practice.

ongoing notesread
active

Technical leadership and decisions

ADRs, ownership, disagreement, escalation, and keeping technical judgment visible.

decision notesread

Engagement log

2024-26Douglas GmbHrewrote the CRM/POS integration in under a year with clearer architecture, more test coverage, and updated documentationConsultant & Technical Lead2022-23Oetker Digital GmbHaudited the codebase and improved quality across the stack: Next.js, TypeScript, GraphQL, Apollo, Cypress, Storybook, TailwindSolutions Architect2022UXMA GmbH & Co. KGdaily pair-programming with two juniors, both grew into independent contributors within weeksTechnical Lead2021Fielmann AGdaily pair-programming as the main knowledge transfer strategyTechnical Architect2020-21Interhyp AGdoubled as agile coach, introducing practices that noticeably reduced sprint frictionEngineering Consultant2019ProSiebenSat.1 Digital GmbHBuilt and maintained React components powering high-traffic show websites for ProSieben and Sat.1 formats, with GraphQL APIs and analytics tracking throughout.Frontend Engineer2019InstaMotion Retail GmbHBuilt a React + GraphQL web magazine for Germany's largest online car dealer, connected to a headless CMS.Frontend Engineer2018Fineway AGearly use of Contentful as a headless CMSFrontend Engineer

Let's talk

If strategy is clear in meetings but weak in the actual delivery system, we can start with where the thread breaks and see whether I can help reconnect it.

who this helps

Engineering leaders, transformation sponsors, tech leads, product people, and teams who need strategy to survive the path from steering meetings to actual delivery.

where it helps

A team or product area where goals, incentives, process, technical reality, or AI adoption pressure no longer pull in the same direction.

how I work

Embedded in teams, close to the people building the product, and connected enough to leadership to make change possible beyond local heroics.

less useful for

AI programs built around hype, copied rollout playbooks, or transformations that mainly need compliance instead of changed work.

Let's talkA short note about the team, the transformation pressure, and where execution currently breaks is enough.