Validation Engineer - Multi-Agent AI Systems (f/m/x)

You will own the virtual validation of this multi-agent AI system in a cloud-based test environment, developing AI-assisted validation strategies that investigate the behavioural space of interacting agents and uncover failures that conventional testing would miss. You build and operate cloud-based validation environments for a multi-agent system that operates across ADAS, Infotainment and backend-service domains – defining test architectures, scenario pipelines, behavioural oracles and regression suites. So, you design AI-assisted exploration strategies to navigate the semantic search space of the system under test: agent-based probing of behavioral boundaries, adaptive test strategies, and optimization techniques to surface edge cases and failure modes. It is your responsibility to develop coverage models and metrics for non-deterministic multi-agent systems and use them to steer test generation and assess release readiness. You establish validation methodologies suited to stochastic AI behaviour: metamorphic testing, property-based testing, adversarial probing and acceptance criteria grounded in confidence acquired rather than binary pass/fail. You build observability and diagnostics tooling for validation runs – distributed tracing, trajectory replay, automated failure clustering and root-cause triage — so that discovered issues can be understood and communicated to the design team. Furthermore, you evaluate and select the LLMs/SLMs that power your validation agents and test-generation systems, balancing latency, cost and accuracy for large-scale test campaigns. You close the feedback loop between validation findings and system improvement by structuring discovered failures, coverage gaps and behavioural anomalies into actionable knowledge for the design and context-engineering teams.

University degree in Computer Science, AI/ML, Systems Engineering, or a related field. 3–5 years of hands-on experience in validation, verification or quality engineering for complex AI, ML or software-intensive systems. Expert-level programming skills in at least one high-level language (C++, Java, Kotlin or Python). Proficiency in AI-assisted development using coding agents search as Claude Code, GitHub Copilot or OpenCode. Understanding of agentic AI architectures, agent patterns and multi-agent interaction. Ability to apply optimization, search or operational research techniques (genetic/evolutionary algorithms, Bayesian optimisation, Monte Carlo methods) to testing or exploration problems. Experience with automotive software quality processes (e.g. ASPICE, ISO 26262) is a plus. You enjoy working in an international team and are passionate about software quality?

Challenging projects with which we are shaping the mobility of tomorrow together. Wide range of personal and professional development opportunities. Attractive, fair and performance-related costs. High level of job security. Annual special payments search as vacation pay, Christmas bonus, and profit sharing. Flexible working hours including 6 weeks annual leave and overtime compensation. Discounted BMW & MINI conditions. Many other benefits at bmw.jobs/benefits Start date: from now on Type of employment: unlimited Working hours: Full-time At the BMW Group, we place great importance on equal treatment and equal opportunities. recruit Ouring decisions are based on the personality, experience, and skills of the applicants. Learn more here. .