Full Stack Software Engineer - RL environments

Acceler8 Talent; San Francisco (CA); On-site; USD 120,000 - 180,000; Full time

Job summary

Acceler8 Talent is seeking a Software Engineer focused on RL Environments to join a fast-growing applied AI research company in San Francisco. You’ll help design the environments, reward signals, evaluations, and data that influence how frontier models are trained and improved.

You will build RL environments, simulations, and task frameworks, develop reward signals and evaluation systems for RLHF / RLVR, and work with researchers to turn training objectives into production systems using Docker

Qualifications

• Strong reinforcement learning experience and hands-on work building RL environments, reward signals, and model evaluation methods.
• Experience delivering RL production systems and data pipelines in fast-moving teams.
• Familiarity with Docker, Kubernetes, and scalable infra for running experiments at scale.
• Experience in RL environments, AI evaluation, benchmarking, or AI safety oriented organizations is a plus.
• High ownership mindset; can thrive in a startup-like setting with rapid iteration.

Responsibilities

• Build RL environments, simulations, and task frameworks
• Develop reward signals and evaluation systems for RLHF / RLVR
• Analyse model and agent failure modes
• Build synthetic and real-world data pipelines
• Work closely with researchers to turn training objectives into production systems

Skills

Reinforcement learning

Environments

Rewards

Model evaluations

Production systems

RLHF / RLVR

Startup experience

AI research

Tools

Docker

Kubernetes

Python

ML frameworks

Job description

Software Engineer — RL Environments | San Francisco

We’re hiring a Software Engineer focused on RL Environments to join a fast-growing applied AI research company working directly with leading frontier model labs.

You’ll help design the environments, reward signals, evaluations, and data that influence how advanced models are trained and improved.

What you’ll do

• Build RL environments, simulations, and task frameworks
• Develop reward signals and evaluation systems for RLHF / RLVR
• Analyse model and agent failure modes
• Build synthetic and real-world data pipelines
• Work closely with researchers to turn training objectives into production systems

What we’re looking for

• Strong reinforcement learning experience
• Experience building environments, rewards, or model evaluations
• Ability to build and deploy production systems
• Experience with Docker, Kubernetes, or similar infrastructure
• Fast-moving, high-ownership mindset
• Experience at an RL environment, AI evaluation, benchmarking, or AI safety organisation
• High-growth startup, early engineer, or founder experience
• Exceptional technical achievement or standout internships

Primarily targeting candidates with 1–5 years of experience, although exceptional new graduates may be considered.