Infrastructure Software Engineering – Platform & Build

Description

The company was founded in 2022 on the view that the speed of producing useful output would eventually limit how much the most capable AI systems could achieve. It builds chips and systems to run frontier AI models, aiming for thousands of tokens per second while handling very long context lengths at scale.

Today the main constraint on progress is the technical and economic cost of fast inference rather than the usefulness of the largest models. Faster inference could support deep reasoning work in drug discovery, software engineering, materials discovery and any field that depends on solving complex problems.

The Infrastructure team builds, scales and tunes the core platform behind the engineering organisation, including a multi-language Bazel monorepo and CI pipelines. Its work covers hardware design and verification, kernel development and ML compilers, so that every engineer can ship quickly and reliably.

The team combines build system depth with an SRE approach, focusing on reliability, build speed, systematic debugging of CI failures, and clear visibility into how the infrastructure performs.

Infrastructure sits within the Software organisation, which builds the full software stack for the company's AI inference systems: ML compilers, device drivers, firmware, runtimes, ecosystem integrations, ML and compute libraries, developer tooling, simulators and datacenter-scale workload deployment. You will work with hardware, software and research teams on AI hardware problems such as verification and simulation, alongside peers from DeepMind, Apple and Meta.

The company has raised $220M from investors including Founders Fund and Accel.

Responsibilities

• Own CI end to end: build, maintain and debug reproducible multi-language pipelines, and improve CI performance across large compute clusters.
• Build and run infrastructure observability, alerting, runbooks and incident response workflows.
• Manage infrastructure as code, provisioning hybrid cloud and on-prem resources from a single source of truth.
• Grow and maintain the Bazel monorepo as it expands across Python, C++, Rust, SystemVerilog and ML workloads.
• Shape the infrastructure and platform experience for every engineer.

Requirements

• Deep infrastructure and build systems expertise combined with an SRE mindset. You approach fault isolation and root cause analysis systematically and use metrics to improve reliability and developer productivity.
• You handle varied work across ML, compilers, kernel drivers, simulators and hardware verification, and you understand that other engineers will build on what you produce.
• 5+ years in software, platform or infrastructure engineering.
• 5+ years hands-on with CI/CD for large-scale products, including monitoring, performance debugging and resolving pipeline failures at scale.
• 3+ years with build systems, ideally modern cross-language ones focused on correctness, performance and extensibility, such as Bazel, Buck, Pants or Please.
• Experience with infrastructure as code using Terraform, OpenTofu or Pulumi.
• Experience with monitoring and observability tools such as Prometheus, Grafana or similar.
• Working knowledge of DevOps and SRE practice, including SLOs, alert design, incident management and on-call.
• Strong proficiency in at least one systems or scripting language for tooling. Python is preferred; Go or Rust are also relevant. A track record of shipping well-tested, maintainable developer tools that others rely on, such as CLIs, CI integrations, and build and release automation.

Conditions

The company works with technologies subject to UK, US and other international export control regulations. Some roles may need additional eligibility checks, and the company will be open about this throughout hiring.

The company is committed to equal employment opportunity and to building a diverse, inclusive team.

Benefits

• A competitive salary that reflects your experience and the specialist nature of the role.
• Meaningful equity so that everyone shares in the value created.
• Private medical, dental and vision cover, a contributory pension, 25 days' holiday plus bank holidays, and life and critical illness insurance.