AI Engineer — Degree Apprentice

The business

About FMTT

FMTT is an established, profitable live-event ticket trading business. Its home ground is UK and European football, international tournaments and major events, sold through the leading ticket marketplaces — and it is now expanding into the American market, where the scope is simple: any live event that trades profitably, we will work. It runs on data — positions, prices, settlements — and it is run by a small, hands-on team, which means the person who writes the code sits next to the people who use it.

The business is mid-way through replacing spreadsheet operations with its own trading platform: a real-time system for positions, market data, pricing and risk — built in TypeScript on Postgres, ingesting high volumes of operational email and live market prices, and using AI throughout. That build is where this role sits. Fuller detail on the business is shared in conversation.

Purpose

Why this role exists

The ambition is explicit: an AI agent for every repeatable flow of the business — reading documents, matching catalogues, pricing against live market data, listing, tracking fulfilment and settlement, producing the reporting — with the goal of automating around 90% of today's manual day-to-day. The team's time moves up a level: supervising the agents, handling the exceptions they surface, and making the judgment calls machines shouldn't. Parts of this are already running in production; most of it is still to be built, and this role exists to help build it.

We would rather grow an AI engineer inside the business — someone who learns the trade and the engineering together — than hire one later who has to learn the industry from scratch. This role is a genuine investment: a degree, a mentor, production responsibility, and a permanent AI Engineer position at the end of it.

The work

What you'll actually do

Ship production software

• Build and maintain desk features end-to-end in TypeScript — Next.js surfaces, Node services, PostgreSQL — with tests on every change and code review on every commit.
• Work the way the desk works: small commits, migrations never applied ad hoc, every change gated by an automated test run before it deploys.

Engineer the data

• Build importers and pipelines that are idempotent — re-running an import updates, never duplicates — and reconcile to the penny against source records.
• Work on high-volume ingestion: systems that receive and structure operational email and documents in real time.
• Handle market-data feeds: live prices, event catalogues, order flow.

Build the agents

The core of the role. Each business flow gets an agent — built, measured, trusted, then handed the work — and agents are assembled from capabilities you'll build and own:

• Document extraction — turning operational email (confirmations, deliveries, sale notifications) into structured, queryable data, by rule and by model.
• Classification — categorising high volumes of messages so the right agent, or the right person, gets exactly what it needs instantly.
• Entity matching — recognising that the same event listed in two places is one event, not two, without ever merging things that differ.
• Pricing & anomaly signals — market floors, margin checks, arithmetic guards that surface exceptions instead of hiding them.
• Agent orchestration — wiring capabilities into end-to-end flows with LLMs in the loop: deciding what an agent may do alone, what it must queue for a human, and how its work is verified. Automation here never means unchecked — every agent works inside the same guards and audits the humans do.
• AI-assisted engineering as daily practice — the platform itself is built with LLM tooling in the loop; you'll learn to direct these tools well, which is fast becoming the core skill of the field.

Learn the trading floor

• Sit with the team. Understand how inventory is sourced, priced, listed and settled — engineers here know the business they're coding, and the best platform features have come from watching the floor work.

Alignment

How the job maps to the degree

Degree apprenticeships assess knowledge, skills and behaviours against real workplace evidence. FMTT's work maps cleanly onto an AI engineering syllabus — this is the alignment, module by module:

Study

The study commitment is real

Stack

What you'll work with

TypeScriptNode.jsNext.js / ReactPostgreSQLSupabaseSQLGitVercelLinuxThird-party APIsClaude / LLM APIsAgentic tooling

Candidate

Who we're looking for

Required

• Meets the university partner's entry requirements — typically A-levels or equivalent (maths valued), and GCSE English and maths at grade 4/C or above. The provider confirms exact requirements and also assesses on aptitude.
• Right to work in the UK and eligibility for apprenticeship funding.
• Genuine intent to complete a four-year programme — this is a commitment on both sides.

What actually matters to us

• Curiosity — has built, automated, scripted or analysed something, on any scale, without being asked to.
• Numerate and precise — cares whether the number is right, and says so when it isn't.
• Writes clearly. Half of engineering is explaining.
• Finishes things.
• Interest in football or live events helps, but isn't required — caring about the problem is.

Offer

What FMTT offers

• A funded degree — BSc (Hons) through the apprenticeship, with no tuition fees and no student debt, while earning a salary.
• Salary of £16,000 in year one, reviewed annually against progress — and always at or above the legal apprenticeship minimum for your age and stage, which itself rises after year one. Apprentice rates are a floor here, not a target.
• Mentorship — direct access to the owner, day-to-day support from the desk, and an AI-first engineering environment most graduates don't see until years into their career.
• Real responsibility — production code from the first months, with the safety rails (tests, review, guards) that make that possible.
• A destination — on completion, a permanent AI Engineer role on the desk with a market-rate salary review. We are training you to stay, not to leave.
• Hybrid working — split between the Blackburn office and home, agreed around the team's rhythm and your study timetable.
• 28 days’ paid holiday a year including bank holidays, per UK standards.
• All equipment provided.

Trust

Boundaries you'll work within

The business handles commercially sensitive and personal data, and the codebase enforces hard rules about it: sensitive values are never stored in plaintext, data is minimised at the point of entry, and sensitive access is audited. You'll be trained on these rules in week one and work inside them from day one — treat this as a feature of the role. Learning security as practice, on a system that genuinely needs it, is worth more than any module. A confidentiality agreement is part of the contract.

Trajectory

Milestones

Culture

How we work

• Tests with everything. Untested code is not done. The build gates the deploy, every time.
• Honest systems. A number that was never counted renders as an unknown, not a zero. Automated checks verify the arithmetic and surface exceptions — never suppressed to make a report look clean.
• Small steps, recorded. Small commits, decisions written down, one phase at a time.
• The workbook wins. When the system and the source disagree, we assume the system is wrong until proven otherwise — and we prove it.

These habits are the fastest engineering education we know how to give.

Process

Next steps

• An informal conversation about the role and the degree — questions welcome in both directions.
• A short practical exercise: nothing to revise for; we want to see how you think, not what you've memorised.
• Together, shortlist the apprenticeship standard and university partner and apply — FMTT handles the employer side of registration and funding.
• Offer, references, and an agreed start date: contact name · email · phone.