Staff Backend Engineer for AI Training Platforms

Dex; Greater London; On-site; GBP 90,000 - 135,000; Full time

Job summary

Frontier AI Lab in London or Paris seeks a Member of Technical Staff to build the distributed backend powering cutting-edge agent training environments. You will own APIs, runtimes, persistence, and data pipelines, turning ambitious research into production infrastructure.

The role emphasizes hands-on architecture, scalable systems, and close collaboration with AI researchers. On-site work at our London or Paris hubs, with a strong focus on reliability, testing, observability, and deployment

Qualifications

• Experience building distributed, data-intensive backend systems in production on AWS, GCP, or similar.
• Strong hands-on capability in typed languages; experience with TypeScript/Node.js preferred; other languages welcome.
• Proven ability to own ambiguous systems from problem statement to shipped infrastructure.
• Willingness to work on-site in London or Paris.

Responsibilities

• Design and operate core backend services, APIs, and persistence layers for dynamic environments.
• Build distributed workflows and high-throughput pipelines for agent interaction traces.
• Collaborate with AI researchers to translate experimental methods into scalable platform capabilities.
• Drive testing, observability, infrastructure as code, and continuous deployment standards.

Skills

Distributed backend

TypeScript

Node.js

Cloud platforms (AWS/GCP)

Ownership of ambiguity

Hands-on architecture

Go

Rust

Java

C++

Tools

AWS

GCP

CI/CD

Job description

Frontier AI Lab in London or Paris seeks a Member of Technical Staff to build the distributed backend powering cutting-edge agent training environments. You will own APIs, runtimes, persistence, and data pipelines, turning ambitious research into production infrastructure.

The role emphasizes hands-on architecture, scalable systems, and close collaboration with AI researchers. On-site work at our London or Paris hubs, with a strong focus on reliability, testing, observability, and deployment