AI Engineer

Type: Full time, Permanent

Location: Blackfriars, London, 2 days working from our headquarters

Seniority: Mid to Senior (SFIA 4)

SC Eligibility Required: You do not need to already hold Security Clearance, but you must be eligible to undergo the vetting process.

About Us

QuantSpark applies data, analytics, and AI to commercial and societal problems that genuinely matter — for financial services, public sector clients, major retailers, international clients and many well-known UK businesses. Our work sits at the intersection of data science, software engineering, and client delivery. The projects are varied, challenging and hugely impactful - what we build often goes live in weeks. We're a 50-person boutique consultancy headquartered in Central London. Small enough that your judgment shapes outcomes; large enough that you'll work across multiple sectors and problem types in your first year. We invest heavily in our people, give genuine ownership, and expect high engineering standards. That means the opportunity to shape important work, the support to keep learning, try new things, and the trust to do your best work without unnecessary bureaucracy.

The Role

We’re looking for an AI Engineer who has already moved from experimentation into production and wants to go further. You may have started in data science or Machine Learning, but now you’re happiest building reliable systems: shipping LLM-powered applications, designing evaluation frameworks, creating retrieval pipelines, and making sure things keep working when real users, messy data, and changing requirements get involved. At QuantSpark, you'll apply that foundation to real client problems. Some engagements call for a carefully orchestrated AI pipeline; others need a well-chosen low-code tool wired up properly. The skill is knowing which approach fits the problem, making it work reliably, and creating measurable impact for the client.

What You'll Be Doing as an AI Engineer at QuantSpark

• Lead delivery on client engagements. You'll own the technical shape of AI and ML workstreams from scoping through handover — making the build-versus-buy-versus-wire-together call, and being the engineer the client trusts in the room.
• Help us win the next engagement. You'll work alongside our Consulting team on live opportunities — translating client problems into credible technical approaches, running discovery sessions, and shaping proposals that are honest about what AI can and can't do.
• Build automation that lands. From orchestrated AI pipelines to a well-wired Power Automate flow, you'll deliver the solution that actually solves the client's problem.

The Problems You'll Solve

In a typical year you might be:

• Turning a PE-backed business's unstructured operational data — contracts, claims, supplier correspondence — into a usable signal, and building the systems that keep doing it.
• Working with a public sector team on fraud detection or predictive maintenance, with the rigour and traceability those clients need.
• Designing an agentic workflow that compresses a week of manual analyst work into hours, with guardrails the client can trust.
• Replacing a brittle Power Automate flow with something better — or building the right one because that's actually what the problem needs.

What You’ll Build

• Production LLM applications — RAG, retrieval pipelines, prompt-engineered workflows, with the evaluation work that makes them reliable.
• Multi-agent systems — tool use, orchestration, memory, and guardrails that keep autonomy useful.
• Classical ML in production — forecasting, classification, optimisation, because not every problem needs a transformer.
• Low-code and workflow automation — Power Automate, n8n, used deliberately when they're the right tool.
• AI harness tooling — Claude Code, Codex, and the broader AI development toolchain.

Most consultancies make engineers pick a lane, we feel our hybrid approach is a better match for the problems clients bring us.

Who You’ll Work With

You’ll sit within our Engineering team and work day-to-day with Product MLOps practice; pre‑sales or technical proposal experience; familiarity with responsible AI practice.

Why QuantSpark

• Variety of work. You'll work across sectors but see engagements through — most engineers are on one or two engagements at a time.
• Engineering quality is taken seriously. Evaluations, monitoring, code review that catches things before they hit clients.
• Real investment in you. £6,000/year for training, upskilling and conferences — one of the most generous budgets in the market.
• EMI share options to share in the reward as we grow.