AI​/ML Engineer – Model Training & Fine-Tuning; Robotics

Position: AI/ML Engineer – Model Training & Fine-Tuning (Robotics)
Ndimensions Labs is a group of technologists from MIT, the University of Waterloo, and the University of Washington, focused on building the infrastructure and learning systems behind next-generation robotics AI. Our focus is not only on deploying models, but on training, adapting, and scaling them for real-world embodied intelligence.
We are developing agents that learn from large-scale multimodal data and improve through iterative fine-tuning, evaluation, and feedback. This role centers on the model training stack: foundation model adaptation, large-scale training pipelines, alignment, and continual learning for robotics.
Position Summary  We are hiring an AI/ML Engineer focused on model training and fine-tuning for robotics AI systems. You will work at the intersection of ML training, multimodal learning, and robotics data systems. The emphasis is on model quality, adaptation, robustness, and generalization.
This position is dedicated to:
Training multimodal models from scratch or from pretrained checkpoints
Fine-tuning foundation models (vision, language, and vision-language-action) for embodied tasks
Designing scalable training and evaluation pipelines
Improving data quality and alignment methods
Responsibilities   Train and adapt multimodal/foundation models (ViTs, VLMs/VLAs, diffusion policies, world models) for embodied robotics.
Build scalable training and fine-tuning pipelines (distributed multi-GPU/multi-node; parameter-efficient fine-tuning such as LoRA/adapters, distillation, QAT).
Own data and iteration loops: curation, filtering, augmentation, and tight feedback with data and systems teams.
Improve policies via learning methods (imitation, offline RL, RL) and rigorous evaluation/ablations (robustness, generalization, failure modes).

Required Qualifications   Strong background in machine learning and deep learning.
Hands-on experience training or fine-tuning large transformer-based models.

Experience with distributed training (e.g., PyTorch DDP, FSDP, Deep Speed).
Understanding of multimodal learning (vision-language, vision-action, or similar).
Experience working with large datasets and building reproducible ML pipelines.
Strong Python skills and familiarity with ML tooling ecosystems.

Preferred Qualifications (Not Required)   Experience training or fine-tuning models for robotics, embodied AI, or control.
Background in reinforcement learning, imitation learning, or policy optimization.
Publications in top ML or robotics venues.

Location:

Boston, MA or Toronto, Canada
Salary range: $100,000 – $300,000 per year, based on qualifications and experience
Equal Opportunity Statement:
Ndimensions Labs is an equal opportunity employer and welcomes applicants from all backgrounds.
Application Instructions  Apply via Linked In or careers with:
CV
Links to relevant work (papers, repositories, training logs, experiments, demos)

#J-18808-Ljbffr