Lead Research Engineer

Atla is a pioneering AI research company committed to engineering safe, beneficial AI systems with a massive positive impact on humanity's future. As Lead Research Engineer, you'll spearhead frontier research in training large language models and advance AI safety research.

The role involves steering LLMs to become strong evaluators aligned with human preferences, defining and executing ambitious research agendas, and developing comprehensive evaluation frameworks. You'll contribute to leading AI safety conferences and journals while working from our London office.

We're seeking someone with 5+ years of pioneering AI research experience, particularly in LLMs, with publications in top-tier conferences. Deep expertise in training and evaluating language models across GPUs (preferably PyTorch) and experience at elite AI labs like OpenAI, DeepMind, or Anthropic is essential.

The ideal candidate is collaborative, comfortable with startup uncertainty, and unpretentious. You'll join a world-class team backed by Y Combinator, Creandum, and founders of Reddit, Cruise, Rappi, and Instacart.

We offer competitive compensation (£200K - £300K), significant equity, comprehensive benefits, and relocation support through UK Government AI Futures Grants (up to £10,000). This is a unique opportunity to shape the future of AI safety while working with cutting-edge technology and world-class researchers.