Research Engineer

Atla is an innovative AI research and deployment company backed by Y Combinator, Creandum, and prominent founders from Reddit, Cruise, Rappi, and Instacart. Their mission focuses on enabling the safe development of artificial general intelligence through training models that can evaluate and ensure the safety of generative AI systems.

As a Research Engineer at Atla, you'll be at the forefront of developing language models as evaluators and implementing safety guardrails for LLMs. This role combines cutting-edge AI research with practical engineering challenges, requiring expertise in training and evaluating language models, particularly using PyTorch across GPU infrastructure.

The position offers an exceptional compensation package, including a salary range of £100K - £300K GBP (approximately $127K - $381K USD), significant equity as an early joiner, comprehensive benefits, and relocation support including visa sponsorship. The role is based in London and includes potential access to UK Government AI Futures Grants for relocation assistance.

Key responsibilities include steering LLMs towards human preference alignment, developing evaluation frameworks, and contributing to top-tier AI conferences. The ideal candidate will have 3+ years of experience, preferably from elite AI research labs, strong empirical research capabilities, and excellent software engineering skills.

The role requires someone who thrives in a fast-paced startup environment, demonstrates strong collaborative abilities, and maintains a commitment to continuous learning. You'll be joining a diverse team of researchers, engineers, and operational leaders, all united in the mission to advance safe AI development.

This is an extraordinary opportunity to make a significant impact in the field of AI safety while working with a well-funded, mission-driven company at the cutting edge of artificial intelligence research and development.