For more than ten years we worked on international automotive programmes, at companies including Bosch and McLaren, under ASPICE and the industry’s quality standards.
We know the morning after a nightly test run: dozens of failed tests, each one a manual search through reports, bus traces, lab schedules, flash logs and Git history. Hours per failure, repeated by different engineers, and many failures closed as “not reproducible”.
We also know the day before the software release: requirements, test cases, results, builds and changes spread across separate tools, and traceability rebuilt by hand in spreadsheets to prove what was tested and why.
None of this work is creative, and all of it is necessary. It is also exactly what modern AI does well, as long as it understands vehicles and stays inside the company.
So we founded TraceSpeak in 2026 to build it.