Latest ML Architect Jobs

Research Engineer, RL Scaling Science

This role involves designing and running large-scale reinforcement learning experiments to understand scaling effects across model size, compute, and task horizon. The engineer will build benchmarks for long-horizon RL, debug system-level issues that emerge at scale, and translate validated research findings directly into production training pipelines. The position sits at the intersection of research and systems engineering, requiring deep empirical rigor and close collaboration with both research and infrastructure teams.

Anthropic London, United Kingdom £375,000 – £640,000 pa

Staff Software Engineer, AI Reliability Engineering

This role focuses on ensuring the reliability and resilience of large language model serving systems at scale. You'll design service level objectives, build observability tooling, lead incident response, and collaborate across teams to strengthen critical infrastructure paths—from API layers to accelerators. The position demands a systems-thinking engineer passionate about robustness, safety, and cross-team partnership in high-stakes AI environments.

Anthropic London, United Kingdom £325,000 – £390,000 pa

Full-Stack Engineer (Frontend-Leaning) - CreativeMusic

This role involves owning the UI for AI music creation, working closely with research to turn new models into user-friendly products. You'll tackle ambiguous problems, collaborate across functions, and ensure seamless integration of new models.

ElevenLabs United Kingdom

Remote Permanent

[Expression of Interest] Research Engineer / Scientist, Alignment - London

Conduct experimental research on AI safety, focusing on alignment and robustness of advanced systems. Develop and test techniques like AI debate and jailbreak evaluation, contributing to publications and safety policy. Work in a collaborative, interdisciplinary team using machine learning and systems engineering.

Anthropic London, United Kingdom

Hybrid Permanent

Anthropic Fellows Program, AI Safety & Security

This role involves conducting full-time empirical AI safety research over a 4-month period, aligned with Anthropic’s priorities such as scalable oversight, adversarial robustness, and mechanistic interpretability. Fellows will work on a self-directed project with mentorship from senior researchers, aiming to produce a public output like a research paper. The program offers a structured environment with access to compute funding, a shared workspace, and integration into the broader AI safety research community.

Anthropic United Kingdom