Member of Technical Staff, Integration/RL Team (Research Engineer)
Cohere · Remote · Paris · London · New York · Toronto · Montreal ·
COMPENSATION
$205k – $380k
Listed on the employer's job board (Ashby)
- Fully remote (UTC-6 to UTC+1)
- Python, JAX, PyTorch
- Distributed RL focus
- Senior engineering level
- 6 weeks paid vacation
Join the Integration team at Cohere to develop and scale machine learning algorithms for large language model post-training. You will focus on large-scale distributed reinforcement learning, designing experiments and writing production code to enhance the quality and speed of training cycles.
This position blends engineering and science, requiring you to implement new tools that support research efforts while optimizing performance across SFT, offline preference, and RL regimes. You will collaborate with infrastructure and scientific teams to build a robust post-training ecosystem.
What you'll do
- Design and write scalable software for model training
- Develop tools to accelerate research and LLM training
- Optimize post-training algorithms for speed and performance
- Collaborate with infrastructure and scientific teams
- Implement techniques for SFT, offline preference, and RL
- Experiment with ideas on cluster and data infrastructure
What you'll need
- Strong software engineering skills and clean code practices
- Proficiency in Python and ML frameworks like JAX or PyTorch
- Experience debugging large-scale distributed training strategies
- Comfortable working with complex ML codebases
- Ability to collaborate with diverse technical teams
Nice to have
- Experience with Kubernetes and Ray frameworks
- Hands-on experience with scalable post-training phases
- Background in ML, LLM, and RL academic research
Stack
- Python
- JAX
- PyTorch
- XLA/MLIR
- Kubernetes
- Ray
How you'll work
- Remote-friendly with offices in major global cities
- Core hours between UTC−06:00 and UTC+01:00
- No minimum in-office attendance requirement
- $500 home office stipend provided
- Co-working benefit available for remote employees
What you get
- 6 weeks of paid vacation annually
- Full health and dental benefits including mental health
- Weekly lunch stipend in local currency
- 100% parental leave top-up for up to 6 months
- Annual enrichment budget for arts, fitness, and learning
- Retirement savings matching (RRSP/401K/Pension)