← Zurück zur Liste
Stelle

Member of Engineering (Reinforcement Learning)

Poolside
AI / ML Engineer • Remote • Vollzeit • 📍 Remote (EMEA
📄 Originalanzeige ansehen (EN)

Poolside, building frontier coding-focused LLMs on its path to AGI, is hiring for its reinforcement learning team, which improves the reasoning and coding abilities of large language models through RL. This hands-on, end-to-end role spans researching new exploration and training algorithms, designing and scaling up RL environments, and implementing ideas across the stack, with access to thousands of GPUs. The team is fully remote across Europe and North America, meeting in Paris three days each month.

Stack

AI/MLPythonPyTorch

Responsibilities

  • ▹Research and experiment on ways to improve reasoning and code generation for LLMs, owning the full experiment life cycle from idea to experimentation and integration
  • ▹Keep up with the latest research and the state of the art in LLMs, RL and code generation, translating research ideas into clean, reusable codebases others can build on
  • ▹Design, analyze and iterate on data generation and training of LLMs
  • ▹Implement and iterate on RL training pipelines that scale reliably across domains
  • ▹Diagnose training instabilities and failures, debug RL runs and propose mitigation methods
  • ▹Write high-quality, reproducible and maintainable code

Requirements

  • ▹Experience with Large Language Models (LLMs): understanding of the Transformer architecture and scaling laws, mid-training and post-training techniques, experience training reasoning and/or agentic models, hands-on use of LLMs with a sense of their capabilities and limitations
  • ▹Reinforcement Learning experience: solid grasp of RL concepts and familiarity with modern algorithms, experience developing distributed, large-scale RL pipelines from data creation to evaluations
  • ▹Strong machine learning, algorithm and engineering background
  • ▹Experience with distributed training
  • ▹Excellent programming skills in Python
  • ▹Familiarity with a deep learning framework (PyTorch or JAX)
  • ▹Ability to discuss the latest research in sufficient detail and be reasonably opinionated

Nice to have

  • ▹Scientific publications in Reinforcement Learning, LLMs or reasoning models

Soft skills

Intellectual curiosityStrong work ethicCollaborative, low-ego mindsetReasonably opinionated

What we offer

  • ▹Fully remote work and flexible hours
  • ▹37 days/year of vacation and holidays
  • ▹Health insurance allowance for you and dependents
  • ▹16 weeks of flexible, full-pay parental leave
  • ▹Well-being, always-be-learning and home office allowances
  • ▹Company-provided equipment
  • ▹Frequent team get-togethers (monthly 3-day gathering in Paris, yearly off-site)
  • ▹Diverse and inclusive, people-first culture