Machine Learning Tech Lead Engineer, Multimodal models (LLM/VLM)

We are looking for an exceptional deep learning algorithm developer with proven experience in LLMs/VLMs/VLA models to join our growing team under Mobileye’s Algorithms Group. The team is building an innovative multimodal learning framework aimed at improving autonomous driving performance by understanding long-tail cases and providing actionable navigation insights. We’re combining state-of-the-art vision-language models to revolutionize how we train, validate, and scale our autonomous systems.

If you’re passionate about deep learning and engineering high-impact autonomous solutions - this is the place for you.

What will your job look like:

• Lead dataset curation strategy - designing scalable pipelines for multimodal data to drive high-quality training and validation.
• Architect and optimize LLMs, VLMs, and VLA models - transforming scene understanding into reliable driving guidance.
• Own the validation strategy - defining methodologies, metrics, and failure-analysis workflows for robust vision-language-action alignment.
• Make an immediate impact by applying deep expertise to guide model design, methodology, and development priorities from day one.
• Collaborate closely with AV planners, perception teams, and infrastructure engineers to ensure seamless deployment in a real-time ecosystem.
• Influence the technical direction of language-driven autonomy - contributing experience, shaping capabilities, and driving innovation to production.
  
  

• M.Sc. in Deep Learning, Computer Vision, NLP, or a related field (Ph.D. an advantage).
• Proven experience with large multimodal or language models (LLMs/VLMs/VLA models) and their real-world integration.
• At least 7 years of hands-on experience in developing deep learning models.
• Strong programming skills in Python (additional C++ is an advantage).
• Experience with modern DL frameworks (e.g., PyTorch, TensorFlow).