Embodied AI / VLA Research Engineer
Maxinsights
- Santa Clara, United States
- On-site
- Mid Level
- 1+ years
- Posted
Skills
Required
- embodied AI
- robotics
- machine learning
- foundation models
- physical
- computer vision
- deep learning
- modern machine learning methods
- Python
- ML/robotics tooling
- GPU utilization / memory usage / training throughput
- distributed data pipelines
Preferred
- VLA
- training
- multimodal
- deployment
- evaluation
- World Model
- robotic manipulation / embodied AI systems
- multimodal learning
- multimodal fusion models
In this role, you will work on the research, training, optimization, and real-world deployment of embodied AI models, including Vision-Language-Action (VLA) models, World Models, and related robotics foundation models.
You will work across multimodal perception, robot control, long-horizon task planning, large-scale robot data, and model deployment. This is a highly hands-on role that involves both model development and real-world robotic system integration
About Maxinsights
- Experience deploying and debugging machine learning models on physical robot systems
- Ability to bring models from development into real-world robotic environments.
Responsibilities
- Vision-Language-Action (VLA), World Model, and other embodied AI foundation models
- Robotic manipulation
- Multimodal perception and control
- Long-horizon task planning
- Vision-language-action reasoning
Robot-environment interaction
- Develop and optimize models for real-world robotic applications.
- Translate research ideas into practical models and systems that can operate reliably on physical robots.
- Foundation Model Training & Optimization
- Train and optimize embodied AI foundation models using large-scale real-robot datasets and egocentric human demonstration data
- Design approaches for incorporating multimodal inputs such as:
- Tactile sensing
- Proprioception
Other robot and environmental signals
- Develop and evaluate multimodal fusion architectures.
- Conduct model training, evaluation, benchmarking, and performance optimization.
- Analyze model performance and identify opportunities to improve training efficiency, generalization, and real-world performance.
- Robotics Data Pipeline
- Build and improve large-scale robotics data pipelines for model training.
- Data cleaning
- Data filtering
- Resampling
- Data augmentation
Data quality evaluation
- Design scalable data pipelines capable of supporting large volumes of robot and human demonstration data.
- Work closely with data and robotics teams to improve dataset quality and training efficiency.
- Model Deployment & Real-Robot Testing
- Deploy trained models to physical robot systems.
real-world robot debugging, testing, and performance optimization
- Diagnose issues across models, sensors, software, and robotic hardware.
- Iterate between model training and real-world testing to improve system performance.
- Help ensure models operate reliably and consistently in real-world environments.
- Work with large-scale real-world robotics and human demonstration data.
- Gain hands-on experience across the full AI development lifecycle, from data pipelines to real-robot deployment.
- Work in a fast-moving startup environment with significant ownership and technical autonomy.
Requirements
- 1+ years of relevant industry or research experience in machine learning, robotics, computer vision, embodied AI, or a related field.
- Strong understanding of deep learning and modern machine learning methods.
- PyTorch
- Experience training and evaluating machine learning models.
- Strong programming skills in
Python
- and familiarity with relevant ML/robotics tooling.
- Ability to work in a fast-paced startup environment and take ownership of technical problems from research through implementation.
- Strong problem-solving and debugging skills.
- Real-World Robotics Deployment
- Experience troubleshooting and stabilizing robotic systems in production or experimental environments.
Hands-on experience with
- Experience optimizing training performance, GPU utilization, memory usage, or training throughput.
- Experience working with large-scale datasets and distributed data pipelines.
- We are looking for an engineer who is excited about the intersection of foundation models and physical robotics
- Hands-on and comfortable moving between research, coding, experimentation, and real-world robot testing
Preferred
- or similar deep learning frameworks.
- Understanding of multimodal learning, computer vision, robotics, or related areas.
- Multimodal / VLA Models
force, tactile, or other multimodal sensing
- Experience designing or training multimodal fusion models.
- VLA models, including model design, training, evaluation, or real-world applications.
- Experience with robotic manipulation or embodied AI systems.
Large-Scale Distributed Training
- Experience with large-scale distributed model training.
- DDP, DeepSpeed, FSDP, or similar distributed training frameworks.
- Interested in solving problems that cannot be addressed through simulation or software alone.
- Comfortable working with large-scale datasets and modern foundation-model architectures.
- Able to take ownership of a problem from data → training → evaluation → deployment → real-world iteration
- Comfortable working in an early-stage environment where priorities can move quickly.
- Curious about emerging VLA, World Model, and embodied AI research and able to translate new ideas into working systems.
- Why Join MaxInsights?
Benefits
- Vision care
- 401(k)
- PTO (Paid Time Off)
Work directly on
embodied AI and robotics foundation models
Collaborate with teams working at the forefront of robotics and foundation-model development.
