About the role
Responsibilities:
- Lead development of Vision-Language Models, Vision-Language Action systems, and diffusion models for robotics, translating research into production-ready AI.
- Research diffusion models for simulation and defect synthesis.
- Build VLM/VLA systems integrating text, CAD, and sensor data.
- Develop auto-annotation pipelines using active learning and synthetic data.
- Optimize models with INT8 LoRA, distillation, TensorRT, and ONNX.
- Own end-to-end AI research lifecycle and collaborate across teams.
- Mentor engineers and publish technical research.
Requirements:
- 8+ years in AI/Deep Learning R& D; or M. S. /PhD in a relevant field.
- Expertise in DDPM, LDM, ControlNet, CLIP, BLIP-2 LLaVA, and Flamingo.
- Strong Python and PyTorch/JAX experience with distributed training.
- Experience with active learning, pseudo-labelling, and weak supervision.
- Familiarity with TensorRT, ONNX Runtime, CUDA/C++ optimisation.
- Strong mathematical foundation in ML and optimisation.
- Experience with Isaac Sim, ROS 2 Nav2 MoveIt 2 Open3D, edge AI, or robotics perception.