feed

Should you hire an AI engineer?

Jordan Richards
Jordan RichardsCEO & Co-Founder · 9 Jul 2026 · Updated 25 Oct 2026 · 4 min read

Hiring an AI engineer is a useful milestone, but it isn't an AI strategy. The harder work is redesigning how the organisation works, and no single hire can carry that alone. The organisations making the fastest progress are building internal capability while drawing on external expertise. At &above, an embedded team that we hire and manage is one way to get there, and the capability transfers to your people as the work ships.

Over the past year, I've noticed a shift in the conversations we're having with businesses.

A year ago, the question was: Should we invest in AI? Today it's: Should we hire an AI Engineer?

On the surface, that feels like progress - organisations are moving beyond experimentation and investing in internal capability. But I think there's a misconception emerging. Whilst hiring an AI Engineer is an important milestone, it isn't an AI strategy. A hire gives you capacity. A strategy sets out where AI creates genuine value, which workflows change and how the business will govern what gets built. Those decisions belong to the whole organisation, not to one new starter.

Too often, organisations expect a single hire to solve a challenge that's fundamentally organisational. They're expected to evaluate tools, build agents, redesign workflows, train colleagues, influence leadership, establish governance, integrate systems, measure ROI and somehow keep pace with one of the fastest moving technologies we've ever seen.

That's an impossible brief. Not because the individual isn't capable, but because AI transformation has never been a one person job. Evaluating tools and building agents is engineering. Redesigning workflows and training colleagues is change management. Governance, leadership buy-in and ROI sit with the people who run the business. Those are different jobs, and one person can't do all of them well while also keeping up with the technology itself.

We've seen this pattern before. Businesses didn't digitally transform because they hired one developer. They didn't become data driven because they hired one analyst. And they won't become AI native because they hire one AI Engineer.

Technology is only one part of the equation, the harder challenge is redesigning how work happens. That means understanding where AI creates genuine value, rethinking workflows across departments, building governance that enables innovation, rather than restricting it, and creating systems that allow knowledge to be shared instead of trapped within teams.

This is where many organisations unintentionally create a bottleneck. The AI Engineer becomes the centre of every AI initiative - every question flows through them, every experiment depends on them, every new opportunity waits in their backlog. Instead of scaling AI across the organisation, they've centralised it around one individual.

The organisations making the fastest progress are building internal capability while drawing on external expertise. Their AI Engineers focus on embedding AI into day-to-day operations and external partners bring experience gained across dozens of organisations, helping avoid common pitfalls, challenge assumptions and accelerate decisions that would otherwise take months to reach. It's not an either-or decision but a complementary partnership.

Internal teams understand the business better than anyone and external partners bring perspective the business can't develop in isolation. There's also a version of this that removes the trade-off entirely, and it's the model we run at &above: we hire the AI engineers, we manage them, and they work embedded inside your team on your workflows. You get the in-house build you wanted without the recruitment risk or the management load, and the capability transfers to your people as the work ships. The engineers work in your standups and your repositories, so your team learns the system as it's built and can own it when the engagement ends.

It's how an embedded squad of ten ran inside Tesco for two and a half years and took creative compliance reviews from four weeks to days. The team started by proving the AI could do the job, with an evaluation suite that tested Gemini against Tesco Retail Media's real approval rules. Then it built the compliance module directly into the advertising workflow, so the work shipped into a live system instead of being left as an experiment. Most AI pilots work. Few become systems the business can run, because that means integrating with the systems, data and rules a business already runs on, and that takes more than one hire.

Hiring an AI Engineer isn't the finish line. The hire gives you someone to build with, but the organisational work starts once they arrive. For most organisations hiring is the starting point, and you don't have to start it alone.