About Ndimensions Labs

At Ndimensions, we are inventing the infrastructure for next-generation robotics AI systems. To train intelligent agents that can understand and act in the world, we need world-class experimentation, data capture, perception, and simulation systems.

We are looking to hire a new team member to drive forward the vision component of our robotics AI stack. This role emphasizes computer vision, multimodal integration, and control, targeting model-level innovations that unify vision, language, and action into robust, low-latency control loops.

The current ecosystem of Vision-Language Agents (VLAs), Vision Transformers (ViTs), Vision-Language Models (VLMs), and diffusion-based planners is evolving rapidly, with no clear “best” solution. We seek someone with deep expertise in computer vision and multimodal learning who is eager to invent, prototype, and question assumptions in this space.

You will work closely with engineers across AI, robotics, and infrastructure to design and scale perception-driven architectures that bridge high-level reasoning with real-time control.

What You'll Do

• Invent and implement new architectures that combine perception, visual understanding, and multimodal reasoning under tight compute and latency budgets.
• Advance the state of the art in robotic perception pipelines, integrating vision with action policies.
• Explore alternatives to current VLA paradigms (e.g., OpenVLA, RT-2, RT-X), identifying limits of existing models and pushing beyond them.
• Extend and adapt existing ViTs (e.g., DINO, SAM), VLMs (e.g., CLIP, BLIP-2, Llava), and diffusion-based planners (e.g., UniPi, Diffusion Policy) for embodied robotics.
• Deploy and test on real robotic platforms in manipulation, navigation, and multi-agent scenarios.

What We're Looking For

• Deep expertise in computer vision, including training/fine-tuning models, transformer-based models, self-supervised learning, segmentation, object detection, and tracking.
• Strong background in multimodal learning (vision-language, vision-action).
• Experience working with large-scale vision datasets, dataset curation/augmentation for robotics, and evaluation protocols.
• Strong engineering skills in C++ and Python, with demonstrated experience building AI or robotics perception systems.
• Familiarity with simulation tools (Isaac, Mujoco, Habitat) and real-world robotic perception pipelines.

Bonus (Not Required)

• Prior work in robotics andor computer vision labs, autonomous systems, or robot learning environments.
• Publications at top research venues.