This page stays with details I can share and verify. I’ll add the fuller decisions, collaboration and lessons once I have written them from my own experience.
Internal systems, repositories and implementation details have been intentionally omitted.
The work so far
My contributions in this area have ranged from CI/CD modernisation to MCP servers and LLM integrations for internal engineering workflows. The work has focused on repetitive engineering activity and the path from a code change to a production-ready artefact.
A clearer path from code to delivery
I led work moving engineering delivery from Jenkins towards GitLab CI. The change involved more than translating pipeline syntax. It was an opportunity to improve automated workflows, simplify delivery and create a clearer path from code change to production-ready artefact.
Exploring where AI actually helps
I have built MCP servers and AI/LLM integrations for internal engineering workflows. The aim has been to explore how language-model-assisted tooling can connect engineering systems and reduce repetitive work without treating AI as a solution in itself.
Areas of exploration have included:
- improving access to internal tools;
- connecting otherwise separate workflow steps;
- automating repetitive engineering activity;
- making operational and engineering context easier to reach;
- testing where human judgement remains essential.
No private URLs, credentials, repositories or internal system details are included here.
A hackathon prototype
I also participated in a hackathon where an automation concept identified an estimated CHF 2.3 million potential saving.
That number represents an estimated opportunity from the concept—not realised or audited savings.