NBCUniversal is seeking a Senior Reinforcement Learning Engineer with a background in developing virtual training environments. Focus on simulating environments and enhancing agent performance in real-world conditions.
In this pivotal role, you will design and optimize high-fidelity simulation environments for training autonomous agents. Your experience in reinforcement learning will help shape complex reward structures and policies crucial for navigating multi-sensor landscapes. Collaborate with cross-functional teams to specify necessary data and simulation requirements within this innovative media and entertainment company.
Key Responsibilities
- Coordinate with ML and Annotation engineers for project specifications
- Develop and manage 2D/3D simulation environments using Unity or Unreal
- Design intricate reward functions to meet product goals
- Implement RL algorithms like PPO or SAC for high-dimensional spaces
- Address reality gaps with domain adaptation techniques
Requirements
- Master’s or PhD in Robotics, AI, or Computer Science
- Proven experience as an RL Engineer in dynamic environments
- Background in robotics, aerospace, or game development
- Proficiency with Python, Git, and Unix environments
- Familiarity with collaborative tools like Jira or Slack
Leverage your expertise in reinforcement learning and environment design to advance NBCUniversal's creative projects.