AI Research Engineer, Model Training & Adaptation

Matter Inc; Sunnyvale, Northern (CA, KY); Hybrid; USD 180,000 - 240,000; Full time

Job summary

Matter Inc is seeking an exceptional Machine Learning Engineer with a deep focus on modeling to lead the end-to-end lifecycle of core ML models for physical AI—from data requirements to training, debugging and deployment into production.

You will own the model lifecycle, iterate across neural architectures, and define data and evaluation strategies to perform well in the real world. The role demands independence and strong collaboration with research and production teams.

Qualifications

• 5+ years of hands-on industry experience training, debugging and deploying ML models into production.
• Strong understanding of modern foundation ML architectures and model optimization techniques.
• Experience with diffusion models and multi-modal architectures.
• Ability to design robust evaluation metrics and run AB tests in production.
• Familiarity with sim2real gaps and strategies to overcome them.

Responsibilities

• Own the Model Lifecycle: training, debugging, and deployment of ML models in production.
• Iterate across neural architectures and post-training techniques to optimize performance.
• Define data requirements and evaluation strategies for real-world performance.
• Bridge research and production by evaluating latest ML publications and open-source models.
• Drive projects forward with minimal hand-holding and own roadmaps.

Skills

Machine Learning

Model Training

Deployment

Diffusion Models

Multi-modal architectures

RL algorithms

A/B testing

Offline Evaluation

Education

PhD in CS/Robotics

Job description

About us

Our vision is to make manufacturing fully autonomous. Matter is building a fleet of industrial assembly

stations driven by close loop neural policies. We train those policies on synthetic data — millions of clips

generated in simulation — then co-train on a small number of real demos.

About the Role

We are looking for an exceptional Machine Learning Engineer with a deep focus on modeling to join our

fast-paced team. In this role, you will own the end-to-end lifecycle of our core machine learning models for

physical AI—from defining the initial data requirements to training, debugging, and deploying them into

production.

• Own the Model Lifecycle:Training, debugging, and deployment of complex machine learning models in a live production environment.
• Algorithmic proficiency: Algorithmic Expertise: Iterate across various neural architectures and post-training techniques to guarantee high model performance, adapting network structures as required.
• Rapid Prototyping & Iteration: Move quickly to test new architectures and techniques. You will build, break, learn, and iterate at a high velocity.
• Define Data & Evaluation Strategy: Establish strict data requirements for your models and design rigorous evaluation strategies to ensure they perform exceptionally well in the real world. Establish strategies to overcome sim2real gaps that can reutilized across tasks
• Bridge Research and Production: Stay up to date with the latest ML publications, open-source models, and industry techniques. Evaluate trade-offs across model families effectively.
• Operate Independently: Drive projects forward with minimal hand-holding. You are comfortable defining your own roadmap and executing against it.

What We Are Looking For

• Education: BS, MS, or PhD in Computer Science, Computer Graphics, Robotics, or a related field (or equivalent experience).
• Experience: 5+ years of hands-on industry experience specifically focused on training, debugging, and deploying machine learning models into production systems.
• Deep ML Algorithms Expertise: Strong understanding of modern & foundation ML architectures and model optimization algorithms. Hand on experience in diffusion models, multi-modal architectures. Proven ability to design robust evaluation metrics, A/B tests, and offline/online evaluation frameworks.
• Research Literacy: A strong pulse on the current ML landscape. You read papers regularly and know how to implement and adapt state-of-the-art techniques.
• Reinforcement Learning: Hands-on experience with RL algorithms PPO / SAC / GRPO, reward modeling in a production setting.
• Contributions to open-source ML projects or published research in the field.