Forward Deployed Engineer (Applied AI)

About OpenKit

OpenKit is an independent AI consultancy, founded in Cambridge in 2020 by Ibrahim Mizi and Reuben McQueen. We help organisations adopt AI properly: we work out what’s actually worth changing, set up tools that reliably work and train teams to use them, and build bespoke integrations only where they measurably pay.

Much of our work is in regulated, high-stakes environments, so we hold ISO 27001, ISO 9001 and Cyber Essentials. Our clients have included the Department for Education, the City of London, DEFRA and several London councils, alongside private firms in finance, legal, healthcare and manufacturing, across the UK, US and Europe.

Why join

We’re hiring because the work is arriving faster than we can deliver it. In the last year we’ve gone from purely bespoke builds to a repeatable audit-and-transformation model, and the pipeline keeps growing.

You’d be our first senior delivery hire. That means clients of your own and a real say in how we price, hire and run the company, with equity that grows as we do. The work covers a genuine range, a regulated fintech one month and a heritage retail brand the next, so the problems don’t repeat.

We tell clients when AI won’t help them, and it’s a large part of why they keep us. Your first months are spent learning our method alongside the founders on live engagements, then you take on clients of your own. If we keep growing the way the pipeline suggests, this becomes the role the next delivery hires report to.

The role

This is a people-facing engineering role. Most clients don’t know where AI can actually help them, so engagements start with an AI audit: interviews and workshops with leadership and each department, to understand how the business makes its money and where the time actually goes. The client leaves with a roadmap and their first tools already running, and after that we embed with their team and deliver the rest. You’d own the whole engagement. In practice:

• You sit inside a business, talking to everyone from leadership to the teams doing the work, and find the bottlenecks AI can reliably clear.
• You work out which tasks AI can genuinely help with, and you say so plainly when it can’t.
• You turn that into a plan: what to change first, what to leave alone.
• You build it, wiring data sources, AI platforms, integrations and configuration into their live systems and getting something working quickly.
• You agree up front what success looks like, whether that’s hours saved or the worst of the grind gone, then show you hit it.
• You train the people who’ll use it until it sticks, so they’re still using it after you’ve moved on.
• You find the next workflow worth doing and keep supporting the client.

The stack

You work across whatever the client already runs: usually one of the major clouds (AWS, Azure or GCP) and the main model providers (OpenAI, Anthropic, Google, or open-weight models where data has to stay on premises), wired into their existing systems through APIs, MCP servers and connectors. There’s no house stack to push. The job is fitting AI into the client’s reality, including sovereign and on-premise deployments, so you need to pick up new tools and sector knowledge quickly and apply them to live work.

What you’ll need

• Strong full-stack development experience that predates LLMs. We use AI coding tools heavily and expect you to, but they need steering by someone who can build and read the code without them.
• Applied AI knowledge from real work: what current models and tooling can and can’t do, and how to tell whether something genuinely helps the people using it.
• Experience in a delivery-facing role such as consulting, solutions engineering, technical pre-sales or implementation.
• You can stand in front of a sceptical room, lay out a new way of working, and train a department until they’re confident. Clients need to trust you and enjoy working with you.
• You judge your own work by whether a team’s day actually improved. A clever build nobody touches is a failure.
• You can run yourself. Embedded means nobody is watching your calendar, and you don’t need them to.
• You can travel to our central Cambridge office.

Nice to have

• You’ve helped a team change how it works, rather than handing over a tool and leaving.
• Deployments into regulated or security-conscious settings such as the public sector, finance or healthcare.
• Hands-on experience with one of the major clouds and its AI services.
• Familiarity with RAG, agent frameworks and current applied-AI tooling.
• A second language. We work across Europe, and French in particular is useful.

Pay and progression

£65,000 base, with a performance bonus tied to clients renewing and actually using what you build, and equity options as the company grows. There’s a budget for training and conferences, and flexibility in how you work.

How to apply

Email [email protected] with your CV and a short cover note, three paragraphs at most, on why you’re a good fit. Include one thing you’ve built that you’re proud of, with a couple of lines on what it is and what you actually did on it. Use the subject line “FDE application” followed by your name.