Applied AI Engineer

AutomateX Ltd; Guildford; Hybrid; GBP 90,000 - 120,000; Full time

Benefits offered by this job

Equity

Job summary

AutomateX is seeking an Applied AI Engineer to lead the build of an AI tool for rail engineers, taking a working prototype to production within a hybrid UK setting (Surrey, Oxford or London). The role demands end-to-end ownership from design to deployment, with the founder closely involved in product direction and client relationships.

You will own both front-end and back-end work (TypeScript/React and Python/Node), ensure secure, governed integration with client document systems, and define

Qualifications

• Shipped LLM applications to real users.
• Able to own full-stack product from frontend to backend.
• Establish evaluation methods and metrics for trustworthy outputs.
• Experience with enterprise integration and security reviews.

Responsibilities

• Take prototype to production with architecture, build and deployment.
• Design grounded extractions from long engineering documents and ensure traceability.
• Define evaluation sets and metrics for accuracy and consistency.
• Connect the tool to client document systems and SSO, and navigate IT governance.
• Collaborate with engineers to translate requirements into deliverable features.

Skills

LLM apps

Retrieval

Full-stack

TypeScript

React

Python

PostgreSQL

Tools

Microsoft Graph

SharePoint

ProjectWise

Single Sign-On

Job description

Applied AI Engineer Upcoming Hybrid · Surrey, Oxford or London · Right to work in the UK

Take a rail AI tool to production. Own it end to end.

AutomateX is an AI-native solutions firm for UK rail. We help engineering teams win and deliver work by putting automation under the hood of the jobs they do every day. We are bootstrapped and built to stay lean and independent.

Compensation Salary + equity

Level Senior

Start shape Paid engagement on a live build

This role is upcoming. We open roles when the client work that pays for them is contracted, and this one is tied to a programme we expect to start in the coming months. We would much rather meet the right person early than hire in a hurry.

The role.

You would lead the build of an AI tool that helps rail engineers review the design submissions they receive from designers and contractors. A working prototype exists and has been demonstrated to engineering teams. The job is to take it to a production system that a large infrastructure client can deploy, govern and trust. The founder stays close to the product and owns the rail judgement and the client relationship. The engineering would be yours, end to end.

What you would do.

• Take a working prototype to production. Architecture, build and deployment, then iteration with the engineers who use it.
• Build grounded extraction over long engineering documents: design statements, calculations, specifications and, in time, drawings. Every output should trace back to its source.
• Design the evaluations. Test sets, accuracy and consistency measures, so that what we claim is what we have measured.
• Connect the tool to the client’s document systems and sign-in (Microsoft 365, SharePoint, ProjectWise, single sign-on), and see it through their IT security and AI governance review.
• Sit with the engineers who use it. Take their requirements, and explain plainly what the tool can and cannot do.

What we look for: technical.

• You have shipped LLM applications to real users. Retrieval, structured extraction or agents, running in production.
• Full-stack enough to own a product: TypeScript and React at the front, Python or Node behind it, Postgres underneath.
• An evaluation habit. You measure before you claim.
• Experience of enterprise integration and security review, for example Microsoft Graph, document management systems or single sign-on.
• A plus: vision models on technical drawings, knowledge graphs, or work in a document-heavy regulated sector.
• The work centres on building and shipping LLM applications. Training models from scratch plays a small part, if any.

What we look for: character.

This matters as much as the code.

• Patience and focus. Real progress in this sector is earned over long cycles. The right person is energised by the slow, deliberate relationship build, and stays steady through it.
• Curious and genuinely excited by change. Happy to explore things that might not work.
• Client-confident. Comfortable in front of clients and leading a technical discussion.
• Honest. You will say when something needs to pivot. That one matters most.

Sector experience.

A background in civil, rail or infrastructure engineering would be a real advantage, though it is not a requirement. What matters most is curiosity: a genuine appetite to learn how the railway reviews and assures its designs.

Location and eligibility.

• This is a full-time, hybrid role. The founder is based in Guildford, Surrey, and regular in-person time is part of how we work. We are best placed to bring on someone within reach of Surrey, Oxford or London who is comfortable meeting in person from time to time.
• You will need the right to work in the UK.

The package.

• Salary plus equity, calibrated to your capability and experience.
• We are bootstrapped and scale with the contracts we win. If you would value ownership, near-term revenue and independence over a hyperscale sprint, this could suit you well.
• It would often start as a paid, milestone-funded engagement on the live build, so that we can both find out whether it fits before a permanent role.

About the founder.

Bill Guo, FICE CEng MEng. Chartered Civil Engineer and ICE Fellow with a decade on major UK rail programmes, building AutomateX to systemise how the sector works. You would be joining someone with the domain, the relationships and a live pipeline.

AutomateX is an equal opportunities employer. We welcome applications from everyone and make decisions on merit, whatever your background.