AI Quality Engineer - Working Student (f/m/d) .
With our Seed Round raised from Lightspeed and BlueYard and leading manufacturing enterprises relying on our systems, we are now scaling fast.
For our world-class team, we are looking for exceptionally driven students who want real ownership early.
Your RoleTesting conventional software is a solved problem.
Testing AI systems is not.
Classic assertions stop working when a system can answer correctly in five different ways, and the industry is still writing the playbook for what comes next.
You define what good looks like for AI features that run in customer production environments.
Kosmos, our platform, is live at industry leaders such as BMW, Audi, BSH, Stellantis, Still, and TDK Electronics.
The job comes with two main.
The first is evals: building the datasets and scoring methods that turn a vague impression of quality into a number the team can act on.
The second is classic engineering rigor: automated end-to-end tests, regression suites, CI that catches issues while a feature is still a branch, and experiments that show where performance and scalability stand.
The mindset we hire for is adversarial.
Your job is to find the edge cases nobody thought of and the prompt nobody expected, and then to prove exactly what happened and why.
For working students who deliver, this is the fastest path into a full-time engineering role.
Your ResponsibilitiesDesign and run evals for our AI features, including datasets, scoring and regression tracking over timeInvestigate performance and scalability bottlenecks, then design experiments to show the impact of the fixExtend our CI pipeline so quality signals reach engineers earlyDo exploratory and edge case testing on new features before they reach customer environmentsDocument and present findings.