AI Engineer (Agentic)

CoreWeave is The Essential Cloud for AI™. Built for pioneers by pioneers, CoreWeave delivers a platform of technology, tools, and teams that enables innovators to build and scale AI with confidence. Trusted by leading AI labs, startups, and global enterprises, CoreWeave combines superior infrastructure performance with deep technical expertise to accelerate breakthroughs and turn compute into capability. Founded in 2017, CoreWeave became a publicly traded company (Nasdaq: CRWV) in March 2025. Learn more at www.coreweave.com .

We're proud to be a Living Wage accredited Employer.

What You’ll Do:

The Physical AI Engineering Team at CoreWeave is responsible for building and scaling the core platform that powers the world’s most advanced engineering simulation and AI workflows. From evolving distributed architecture to enabling the next generation of agentic applications and physical AI, this team delivers the performant, reliable, and scalable foundation trusted by the world’s largest engineering companies.

About the role:

We're seeking a Senior AI Engineer to build the agentic layer of our Physical AI platform. You'll design, build and ship the agents behind our engineering copilot: systems that take a real engineering problem from an automotive, industrial or robotics team, explore the data, build and evaluate models, and return results an engineer can trust and act on.

This is a hands-on engineering role, not a research role. You'll work alongside our researchers and software engineers: Research explores new approaches, and you turn the ones that work into reliable production systems. You'll own agents end to end – from tool and context design, through evaluation, to deployment and operation in production.

In this role, you will:

Agentic Systems

• Design and build agents that run the auto-research loop on customer engineering problems: data exploration, feature and model iteration, evaluation and reporting.

• Define tool interfaces and MCP servers that expose platform capabilities (data connectors, Physical AI libraries, experiment tracking and model registry, compute) to agents in a safe and predictable way.

• Own context and memory design for long-running agent tasks, including retrieval over engineering data, documentation and past experiments.

• Select and integrate models, balancing quality, latency and cost.

Evaluation & Reliability

• Build evaluation harnesses and benchmarks that measure agent performance on real engineering use cases, so quality is measurable and regressions are visible.

• Instrument agents with tracing and observability and use that data to improve them.

• Design guardrails, sandboxed code execution, permission models and human-in-the-loop checkpoints so agents act safely on customer data and compute.

• Take agents from prototype to production: testing, CI/CD, deployment, monitoring and on-call ownership.

ML & Domain Work

• Contribute to ML work alongside the ML Engineers – model training, evaluation and Physical AI library development – especially where agents use or generate these components.

• Translate the vocabulary and workflows of physical engineering domains into agent behaviour, working with Forward Deployed Engineers and customers.

• Cover ML engineering work when needed, and help ML Engineers adopt agentic tooling in their own workflows.

Collaboration

• Work with Research to turn promising agentic approaches into production systems.

• Partner with platform engineers on the agent runtime, workflow orchestration and auth as the platform moves to a distributed, agentic-ready architecture.

• Work with Frontend and Product to expose agent capabilities in the product in a way non-ML engineers trust.

• Mentor engineers, and interview and onboard new team members.

Who You Are:

• 5+ years of professional software engineering experience, including 2+ years building production systems on top of LLMs. Your background is in shipping and operating software, not academic research.

• You have shipped at least one agentic system to production with real users: tool-using agents, multi-step workflows or code-generating agents.

• Strong production Python.

• Hands-on experience with LLM APIs and agent frameworks, and a clear view of where frameworks help and where they get in the way.

• Experience designing tools and APIs for agents, including MCP or equivalent tool-calling protocols.

• Experience building evaluation for LLM or agent systems: datasets, graders, LLM-as-judge, regression tracking.

• Solid ML fundamentals. You can train and evaluate a model with scikit-learn or PyTorch, and you understand validation, overfitting and uncertainty well enough to judge an agent's output.

• Experience with retrieval and context engineering for agents.

• Strong software engineering practices: code review, automated testing, CI/CD, observability, Docker and Kubernetes, and at least one major cloud (AWS preferred).

• Security-minded: sandboxing, authentication/authorisation, prompt injection and data isolation in multi-tenant systems.

• Clear technical communication with both engineers and non-ML stakeholders.

Preferred:

• Experience with auto-research, data science agents or agents that write and run experiments (e.g. in notebooks).

• Experience with tabular, time-series, sensor, test or simulation data (automotive, aerospace, industrial or robotics).

• Previous experience as a backend or platform engineer before moving into AI engineering.

• Familiarity with experiment tracking tools and notebook environments.

• Experience serving open-source models on GPU infrastructure.

• Experience with workflow orchestration for long-running agent tasks.

• Experience with fine-tuning or reinforcement learning to improve agent behaviour.

Wondering if you're a good fit?…

Source: Arbeitnow (https://www.arbeitnow.co.uk/jobs/companies/coreweaveu/ai-engineer-agentic-london-473164)