AI Harness Engineers

Capgemini; Greater London; On-site; GBP 90,000 - 130,000; Full time

Benefits offered by this job

Disability Confident Employer

Job summary

Capgemini seeks a senior developer engineer to own the end-to-end AI agent development loop, from environments to harnesses and evaluation gates.

You will instrument, optimize, and scale CI/CD pipelines, with a focus on reliable agent-based code execution and rapid feedback cycles across human and agent teams.

Qualifications

• Prior ownership of a development environment, build system, or paved-path workflow used by a multi-team engineering organisation.
• Strong Python plus container and Kubernetes fluency; comfort operating CI/CD systems at scale.
• Direct experience deploying or operating AI coding agents (Claude Code, Cursor, Copilot, or in-house), beyond personal use.
• You follow frontier agentic-systems research (harness design, reinforcement learning from execution feedback, evaluation methods) closely enough to put it into production within the quarter it lands.
• A measurement habit: you can show numbers for a developer-experience improvement you shipped.
• Daily, hands-on use of AI coding assistants as part of your own development workflow.
• Hermetic build systems (Bazel, Buck, Nix, or similar) or monorepo tooling at scale.
• Go or Rust; Git-at-scale experience.
• You have built one-shot or unattended agent pipelines with hard failure caps and human escalation.
• You have turned agent execution traces into training or evaluation datasets, or built reinforcement learning pipelines from execution feedback.
• You have written publicly about developer experience or agent harnesses.
• Financial services engineering exposure (banks, insurers, or payment providers).

Responsibilities

• Development environments end to end: fast, isolated, reproducible, for human engineers and for agent fleets (sandboxes, eternal environments, warm starts)
• Deterministic CI and the pre-push validation surface, so failures are caught at the desk, not in the pipeline
• The agent harness: the tooling, permissions, retry limits, and orchestration blueprints our coding agents operate within, improved permanently every time an agent fails
• The recursive improvement loop: agent transcripts, failure modes, and evaluation verdicts flow back into harness changes and model adaptation datasets automatically, so the system that builds our products improves itself
• Evaluation-gated merges: the CI integration that makes eval suites a first-class merge gate
• Measurement: cold-start times, agent PR merge rates, review-time economics; you improve what you instrument

Skills

Python

Kubernetes

CI/CD

Go

Rust

Git at scale

AI agents

Tools

Bazel

Buck

Nix

Job description

About the job you’re considering

Hybrid working:

The places that you work from day to day will vary according to your role, your needs, and those of the business; it will be a blend of Company offices, client sites, and your home; noting that you will be unable to work at home 100% of the time.

If you are successfully offered this position, you will go through a series of pre-employment checks, including, identity, nationality (single or dual) or immigration status, employment history going back 3 continuous years, and unspent criminal record check (known as Disclosure and Barring Service)

About us

We build products, not projects: software for insurance claims, payment operations, and health operations, sold to banks, insurers, and health plans. Three product lines run on one shared platform, built by a deliberately small, senior team. Our engineering model is agentic: engineers author the specifications, tooling, evaluation suites, and guardrails, and AI agents do most of the implementation. Humans own every consequential decision, and in our regulated domains some decisions are human-only by design.

The role

Our engineering model depends on infrastructure most companies treat as an afterthought. AI agents implement most of our code, which means the development loop (environments, builds, CI, the harnesses agents run in) is the engine of the whole organisation's velocity. Every minute an engineer or an agent waits on an environment, a flaky test, or an unreviewable diff is product not shipped. You will own that loop end to end, for people and for agents. The harness is a self-improving system: every failure, transcript, and evaluation verdict is fuel for the next version of the harness and, increasingly, for the models inside it.

What you will own

• Development environments end to end: fast, isolated, reproducible, for human engineers and for agent fleets (sandboxes, eternal environments, warm starts)
• Deterministic CI and the pre-push validation surface, so failures are caught at the desk, not in the pipeline
• The agent harness: the tooling, permissions, retry limits, and orchestration blueprints our coding agents operate within, improved permanently every time an agent fails
• The recursive improvement loop: agent transcripts, failure modes, and evaluation verdicts flow back into harness changes and model adaptation datasets automatically, so the system that builds our products improves itself
• Evaluation-gated merges: the CI integration that makes eval suites a first-class merge gate
• Measurement: cold-start times, agent PR merge rates, review-time economics; you improve what you instrument

What you will need

• Prior ownership of a development environment, build system, or paved-path workflow used by a multi-team engineering organisation
• Strong Python plus container and Kubernetes fluency; comfort operating CI/CD systems at scale
• Direct experience deploying or operating AI coding agents (Claude Code, Cursor, Copilot, or in-house), beyond personal use
• You follow frontier agentic-systems research (harness design, reinforcement learning from execution feedback, evaluation methods) closely enough to put it into production within the quarter it lands
• A measurement habit: you can show numbers for a developer-experience improvement you shipped
• Daily, hands-on use of AI coding assistants as part of your own development workflow
• What sets you apart
• Hermetic build systems (Bazel, Buck, Nix, or similar) or monorepo tooling at scale
• Go or Rust; Git-at-scale experience
• You have built one-shot or unattended agent pipelines with hard failure caps and human escalation
• You have turned agent execution traces into training or evaluation datasets, or built reinforcement learning pipelines from execution feedback
• You have written publicly about developer experience or agent harnesses
• Financial services engineering exposure (banks, insurers, or payment providers)

The reference stack

The reference technology stack for this role is our supported paved road: self-hosted LangSmith and LangGraph Platform as the agent runtime and evaluation plane, model providers behind a swappable gateway seam, PostgreSQL with pgvector plus ClickHouse and S3-compatible object storage as the data platform, Neo4j Enterprise as the semantic knowledge graph, an agent memory plane serving episodic and precedent memory over MCP, MCP-native connectors, OpenTelemetry and Grafana for observability, all on CNCF-conformant Kubernetes with Helm and Argo CD, deployable to any hyperscaler or on-prem. A tool-for-tool match is not expected: analogous experience counts fully. If you have built and operated systems of this shape on comparable components (a different orchestration framework, graph engine, evaluation platform, or serving stack), you have what we are looking for.

How we work

Engineers write specs, harnesses, evals, and guardrails; AI agents execute the implementation loops. Review, not typing, is where engineering judgment goes.

Three human gates govern everything we ship: spec approval, merge, and release. Regulated code paths (money movement, authentication, cryptography, secrets) are always human-owned.

Small and senior by design. No separate QA function, no scrum masters; quality comes from evaluation gates and whole-team review rituals.

Domain experts (claims practitioners, payment scheme experts, clinicians) are full-time members of the product teams you will serve.

Success in year one

• Fresh environment cold-start is measured in seconds, for humans and agents, and you can prove it
• Agent-written PRs merge at a rising rate with flat or falling human review minutes per feature
• Every recurring agent failure mode from the last two quarters has a permanent harness fix, and the harness has also lost components that model progress made unnecessary

We are a Disability Confident Employer

Capgemini is proud to be a Disability Confident Employer (Level 2) under the UK Government's Disability Confident scheme.As part of our commitment to inclusive recruitment, we will offer an interview to all candidates who:

• Declare they have a disability, and
• Meet the minimum essential criteria for the role.

Please opt in during the application process.

Make it real - what does it mean for you?

• We realise a Total Reward package should be more than just compensation. At Capgemini we offer range of core and flexible benefits and have a Peer Recognition Portal called Applaud.
  To help support wellbeing we have trained 'Mental Health Champions' across each of our business areas, and we have invested in wellbeing apps such as Thrive and Peppy.
• You will be empowered to explore, innovate, and progress. You will benefit from Capgemini’s 'learning for life' mindset, meaning you will have countless training and development opportunities from thinktanks to hackathons, and access to 250,000 courses with numerous external certifications from AWS, Microsoft, Harvard ManageMentor, Cybersecurity qualifications and much more.

Capgemini. Make it real.

Why you should consider Capgemini

Growing clients’ businesses while building a more sustainable, more inclusive future is a tough ask. When you join Capgemini, you’ll join a thriving company and become part of a collective of free-thinkers, entrepreneurs and industry experts. We find new ways technology can help us reimagine what’s possible. It’s why, together, we seek out opportunities that will transform the world’s leading businesses, and it’s how you’ll gain the experiences and connections you need to shape your future. By learning from each other every day, sharing knowledge, and always pushing yourself to do better, you’ll build the skills you want. You’ll use your skills to help our clients leverage technology to innovate and grow their business. So, it might not always be easy, but making the world a better place rarely is.

About Capgemini

Capgemini is an AI-powered global business and technology transformation partner, delivering tangible business value. We imagine the future of organisations and make it real with AI, technology and people. With our strong heritage of nearly 60 years, we are a responsible and diverse group of 420,000 team members in more than 50 countries. We deliver end-to-end services and solutions with our deep industry expertise and strong partner ecosystem, leveraging our capabilities across strategy, technology, design, engineering and business operations. The Group reported 2024 global revenues of €22.1 billion.

Make it real