Machine Learning Engineer (m/f/d) VLA Training & Evaluation

About the role

<p><strong>The AI Teams at Agile Robots</strong> are looking for a <strong>Machine Learning Engineer (m/f/d) VLA Training &amp; Evaluation</strong>, who will build the frameworks for fine-tuning and benchmarking robot foundation models on collected robot data. The frameworks are meant to work across different foundation models and for teams with modest data and compute.</p> <h3><strong>Responsibilities</strong></h3> <ul> <li><strong>Training Workflows:</strong> Build reusable, model-agnostic training and evaluation workflows, with experiment tracking, model versioning, and checkpoint management, so that different foundation models can be swapped in without rebuilding the pipeline.</li> <li><strong>Fine-Tuning:</strong> Implement fine-tuning of pretrained vision-language-action models on collected robot data, for example with parameter-efficient methods such as LoRA or adapters, or with distillation.</li> <li><strong>Policy Training:</strong> Build training loops for action-chunking and transformer-based action models, diffusion policies, and VLM/VLA heads.</li> <li><strong>Benchmarks:</strong> Build benchmarks with generalization splits, success metrics, and failure-analysis templates.</li> <li><strong>Transfer Studies:</strong> Work with simulation engineers on studies, including metrics and ablations, that test whether improvements in simulation realism improve performance on the real robot.</li> <li><strong>Research Baselines:</strong> Turn published methods into working baselines and templates that others can extend.</li> </ul> <h3><strong>Essential Skills</strong></h3> <ul> <li><strong>Academic Background:</strong> Degree in Computer Science, Machine Learning, Robotics, or a related field (Master’s or PhD typical), or an equivalent combination of formal training and demonstrated experience training deep learning models.</li> <li><strong>Training Internals:</strong> Hands-on experience writing and debugging training loops in PyTorch or JAX, including data loading, loss design, and optimizer behavior.</li> <li><strong>Fine-Tuning:</strong> Experience fine-tuning pretrained models and adapting them to new tasks and data.</li> <li><strong>Robot Learning:</strong> Working understanding of how imitation-learned policies, such as action-chunking or diffusion policies, turn observations into robot actions.</li> <li><strong>Experiment Management:</strong> Experience designing evaluations and ablations, and running experiments with tracking, versioning, and checkpoint management so that others can reproduce them.</li> </ul> <h3><strong>Beneficial Skills</strong></h3> <ul> <li><strong>VLA Models:</strong> Experience fine-tuning vision-language-action or vision-language models for robot control.</li> <li><strong>Real-Robot Evaluation:</strong> Experience evaluating learned policies on physical robots.</li> <li><strong>Sim-to-Real:</strong> Familiarity with measuring how well models trained in simulation transfer to real hardware.</li> </ul><div class="content-conclusion"><h3 class="TranslationField-module__fieldItemName___Q0F_7 TranslationField-module__aboutUsWrapper___WXJpq"><strong>What we offer</strong></h3> <ul> <li> <div>A dynamic high-tech company, combined with financial soundness and world-class investors.</div> </li> <li> <div>Join an interdisciplinary, international team with 60+ different nationalities in a collaborative work environment.</div> </li> <li> <div>Lots of development opportunities as we continue to grow.</div> </li> <li> <div>Challenging tasks and impactful projects alongside experts that enable professional and personal growth.</div> </li> <li> <div>Corporate Benefits Program covering health, mobility, and learning for 100€ net per month.</div> </li> <li> <div>Modern office facilities with a rooftop terrace overlooking Munich, free drinks &amp; fruits, and regular company events contribute to a good working environment</div> </li> </ul> <h3><strong>About us</strong></h3> <div class="TranslationField-module__f

Skills