I help engineering leaders build AI-assisted systems they can explain, defend, and stand behind. I've worked in software testing since 2005, and I still build hands-on every day. What I do now is help teams keep accountability in step with how fast AI lets them move.
2026 speaking: Keynote + workshop at HUSTEF (Budapest, October), talk at Quality Beacon (Copenhagen, October), talk at Øredev (Malmö, November), and talk at Agile Testing Days (Potsdam, November). All on AI accountability and what it actually takes to test AI-assisted systems.
CodePath: Senior test developer and release manager on the backend team of a US education nonprofit, a long-running client engagement of Northern Test Consulting since 2021.
The advisory work is grounded in real systems. I build this AI-assisted infrastructure myself, and these are two of the tools that come out of it.
A hands-free session recorder for exploratory testing. You test in the browser and narrate your observations out loud. The tool captures your actions, transcribes your voice, and uses an LLM to structure your findings into bug reports, without ever deciding what is a finding. That's your job.
Built on a simple principle: AI should assist, not decide. The tester drives intent, controls meaning, and owns judgment. The tool just listens and writes the notes.
A production-grade personal AI stack built on a single principle: AI should assist, not decide. A semantic knowledge base with hybrid search and a Data Constitution. A personal assistant that drafts but never sends without approval. An evaluation framework with sandboxed fixtures and holdout sets.
Six interconnected systems, 19 automated services, 8 data sources. Built by a testing practitioner, tested like production software.
Based in Karlskrona, Sweden. Available for advisory engagements, speaking, and workshops.