AI Delivery Architect in Slough

Level: Senior — engineer-turned-architect, still hands-on

Type: Permanent · Location: London (hybrid, some client travel)

Reports to: Delivery Lead

About Lumo

Lumo is an AI transformation consultancy. We help businesses embed AI into their SDLC, processes, and practices — from advisory through to hands-on application and data build. Our delivery model is AI-native: small senior teams orchestrating AI tooling (Claude, Claude Code, Devin, Cursor/Windsurf, agentic workflows) to ship work faster, leaner, and with more rigour than traditional consultancies.

Why this role exists

This role is built around a strong, experienced engineer-turned-architect — someone with real delivery behind them and the credibility that comes with it. You do not need to be an AI expert today. If you have deep AI tooling experience, brilliant. If you don't, but you have the engineering depth and a genuine track record of learning fast, we will back you to pick it up here. The experience is the hard part; the AI we can teach.

Here's the context. Our clients are trying to change how they build software — they want AI woven through their delivery lifecycle, not as a novelty, but as the default way their teams design, build, test, and ship. They need someone credible who can sit alongside their engineering and product leaders, work out what that actually looks like for them, and help them get there. Credibility is everything: you have shipped real systems, you have led delivery, and people listen to you because of it.

That is this role. You work directly with client teams in an advisory capacity, helping them evolve their AI SDLC — where AI tooling earns its keep, how delivery changes around it, what good looks like, and how to get there without breaking what already works. It is not a whiteboard-only solution architect role: you stay hands-on, with AI tooling doing serious work alongside you, because the point is to demonstrate the way of working, not just describe it. Whether you arrive already fluent in those tools or grow into them with us, the foundation is the same — proven engineering judgement and the appetite to keep learning.

What you'll do:

• Work with client teams. Sit alongside our clients' engineers, architects, and product leads to understand how they deliver today and where AI changes the picture. Run working sessions, advise on practice and tooling, and translate the work for technical and non-technical stakeholders without losing precision.
• Help evolve their AI SDLC. Shape how AI tooling fits into a client's delivery lifecycle — design, build, review, test, release. Recommend where to lean in, where to be cautious, and what good looks like, then help them adopt it.
• Stay hands-on with AI. Show, don't just tell. Use Claude Code, Devin, Cursor/Windsurf, and agentic workflows as a force multiplier. Know what to delegate to agents, what to keep yourself, what to put behind an eval, and what to throw away.
• Set credible technical direction. Bring architecture judgement built on real delivery — sensible system design, integration patterns, and trade-offs that hold up in production, not just on a slide.
• Partner with the Lumo delivery team. Work with our delivery lead and the wider team to scope engagements, raise risks early, and keep things moving. We are a start-up, so the shape of the team is still forming — you are comfortable with that.

What we're looking for

Must-have

• A strong engineering background, with a backend bias. You came up building production systems — services, data, integration — and you have a deep appreciation for how technology actually gets delivered.
• A move into architecture, while staying hands-on. You took on architecture and led deliveries, but you never stopped building. You are not a diagrams-only architect.
• A track record of learning and adapting. This matters most. You can point to times you picked up an unfamiliar stack, tool, or way of working and shipped with it. You are self-sufficient and you thrive in new delivery methods rather than waiting to be shown.
• Credibility with technical leaders. You have led delivery and can hold your own in a room with a client's senior engineers — advising, challenging, and being trusted.
• Comfortable working directly with clients. You can run a working session, explain progress and trade-offs clearly, raise risks early, and adjust register for technical and non-technical audiences.
• Enough breadth to use AI across the stack. You don't need to be a specialist front-end engineer or platform engineer, but you understand front-end, IaC, and the broader stack well enough to direct AI tooling across them confidently.
• London-based and happy with hybrid working and occasional travel to client sites.

Nice-to-have (and genuinely fine to learn on the job)

• Daily, sophisticated use of AI coding tools (Claude Code, Devin, Cursor/Windsurf, Copilot) with informed views on where each earns its keep.
• LLM application engineering — agent orchestration, RAG, structured outputs, evals.
• Prior consultancy or client-facing delivery experience.
• Cloud and data infrastructure depth (GCP/AWS, Terraform, CI/CD, warehouses, ETL).
• Front-end or IaC delivery experience to complement the backend core.

How you'll know this job is for you

You like being the person clients trust to tell them the truth about how to build better — and then to roll up your sleeves and show them. You would rather learn a new tool and ship with it than defend the way you have always worked. And you are energised, not unsettled, by a start-up where the role will keep changing as AI delivery itself changes.

How we work

AI-native delivery is the default. We are a small senior team because we orchestrate AI tooling well. If you would rather pretend that isn't happening, this isn't the place.

Honest, not over-sold. Our client posture is candid. We flag risk, recommend gates, and prefer documentation and handover over perpetual retainer. You'll be expected to do the same.

Fast paced. Change is relentless. Roles, teams, and methods are evolving — that is the opportunity, and it is why self-sufficiency matters so much.