AI is being deployed in places where failure isn't an option - defence, healthcare, public safety, critical infrastructure. We build the testing software that lets organisations verify these systems actually work: their performance under stress, their robustness, their safety.
Our clients include government agencies, military organisations, and commercial operators running AI in high-stakes environments. We work through embedded teams deployed on-site or closely integrated with our clients. We were founded by Temasek, are headquartered in Munich and Singapore, and are currently scaling across Europe.
The role
You’ll build models, model-based features, and the AI capabilities inside our product. This role is implementation-led but research is part of the job – you’ll read papers, prototype, and ship.
The work spans model development, evaluation methodology, synthetic data, MLOps, and agentic systems.
Depending on your strengths, you may weight toward LLM, computer vision, reinforcement learning, synthetic data, agentic AI, or robotics/edge AI. We expect depth in your primary area plus working knowledge across the others.
Core: Python, PyTorch or TensorFlow, model serving and MLOps tooling, Git
Also in play: CrewAI, LangGraph, LlamaIndex, MCP, ROS2, Gazebo/Isaac Sim, edge inference stacks
What you’ll work on
- Building, deploying, and maintaining AI models in production – from fine-tuning through serving and monitoring
- Designing evaluation methodologies, test plans, and quality frameworks for LLM, CV, and RL systems
- Building agentic systems using frameworks like CrewAI, LangGraph, LlamaIndex, and MCP
- Creating synthetic data pipelines, adversarial test cases, and benchmark datasets
- Owning model serving infrastructure and deployment automation, including edge and air-gapped customer environments
- Tracking SOTA across your domain and translating findings into product improvements
What we’re looking for
- 3–7 years in AI/ML engineering with production deployment experience (not just notebooks or research code)
- Strong Python and hands-on experience with PyTorch, TensorFlow, or similar
- Demonstrated depth in at least one of: LLM/NLP, computer vision, reinforcement learning, synthetic data generation
- Experience taking models to production: serving, API integration, monitoring, versioning
- Sound grasp of evaluation methodology – experimental design, statistical rigour, benchmark creation
- Able to bridge research and engineering: read papers, prototype, ship production code
- Clear communicator who can explain AI concepts to non-specialist colleagues and customers
Bonus
- MSc or PhD in AI/ML, Computer Science, or related field
- LLM fine-tuning, RLHF, or production-scale prompt engineering
- Reinforcement learning: reward modelling, policy evaluation
- ROS2, edge AI, or simulation (Gazebo/Isaac Sim) for robotics
- Published research or open-source contributions
- Experience in defence, pharma, or other regulated industries
- Air-gapped or on-premise deployment experience
- Munich-based, hybrid Tue‑Thu in office
- Working language is English; German is a plus
- You’ll join a scaling company with co-HQs in Munich and Singapore, growing the engineering org significantly over the next 18 months
If this sounds like your kind of problem, we’d like to hear from you.
