Job Description

Employment Type: Full-time

Position Overview

The Smart Agents group is responsible for building Machine Learning (ML) and Reinforcement Learning (RL) models and systems to simulate road users in various situations. These simulations generate scenarios used for testing and training Autonomous Vehicle (AV) driving policies. The team develops ML/AI models that control other characters (vehicles, bikers, pedestrians) in simulation environments to interact realistically with AVs. The technology stack includes Generative AI models (GPT) and RL policies. The Smart Agents group collaborates closely with Simulation, Behaviors, Perception, and Safety Engineering teams.

Remote: This role is based remotely. However, if you live within a 50-mile radius of Atlanta, Austin, Detroit, Warren, or Mountain View, you are expected to report to that location at least three times a week.

Responsibilities

• Support the team in developing ML and RL models, including training loop development and optimization.
• Streamline integration and create ML infrastructure, metrics, and data pipelines for production model deployment and rapid experimentation.
• Contribute strong software engineering (SWE) expertise as part of the ML team.
• Assist the ML team in accelerating project timelines, particularly in areas related to Autopilot, Lane Keep, and AV technologies.
• Develop ML/RL models and training loops.
• Optimize training loops, streamlining integration, and creating ML infrastructure, metrics, and data pipelines for production model deployment and fast experimentation cycles.

Requirements & Skills

• 4+ years of experience in the field of robotics or latency-sensitive backend services.
• Background working with machine learning teams, algorithms, and models.
• Proven experience in machine learning and classification.
• Familiarity with ML frameworks such as TensorFlow or PyTorch.
• Strong programming skills in modern C++ or Python.
• Expertise in setting architectures that are scalable, efficient, fault-tolerant, and easily extensible.
• Ability to design across multiple systems, investigate sophisticated areas, and possess a broad understanding of systems outside of your domain.
• Ability to wear multiple hats, shifting between coding, design, technical strategy, and mentorship, with excellent judgment on when to switch contexts.
• Track record in deploying perception/prediction/AV models into real-world environments.
• Passionate about self-driving car technology and its impact on the world.

Bonus Skills:

• Experience building highly performant ML and system pipelines.
• Experience with profiling CPU and/or GPU software, process scheduling, and prioritization.

Desired Background

Experience in simulation and robotics is highly desirable, with a preference for candidates from AV or robotics backgrounds rather than solely cloud-focused companies.