From marketing operations to AI engineering: what my placement year actually taught me

The clearest example was a traffic problem. More people were getting answers straight from AI tools instead of clicking through to a website, and Ancestry's organic search traffic had dropped by close to 30% because of it. I ended up building the strategy to recover it, which meant learning how large language models actually retrieve and rank information well enough to work backwards from it. That's a fairly technical problem to land in your lap during a marketing placement, and it's the point where I stopped thinking of AI as something adjacent to my role and started thinking of it as the role.

The rest of the placement followed the same pattern. I built an automation workflow that cut a chunk of manual reporting by around 20%, ran a training session for 20-plus leaders from other teams on using AI day-to-day, and piloted an integration connecting Claude to some of our internal systems, mostly to see if it would help. None of this was in my original job description. I just kept noticing problems that automation could solve and built the solution instead of waiting to be asked.

That's the part of my placement year I'd tell any Kingston University student to pay attention to: your job title is the floor, not the ceiling. The tasks nobody assigned you are often where you learn the most about what you're actually good at.

By the time I started applying for graduate roles, I wasn't trying to become "an AI person" from nothing. I was looking for something that matched the direction I'd already drifted in. That meant reaching out directly to people already doing the kind of work I wanted, tailoring every application instead of sending the same CV everywhere, and treating every rejection as information about what I needed to show better next time. I got to second and third interview rounds more than once before hearing no. It still took months.

The one that did work out, at Janus Henderson Investors, didn't exist in its current form when I first spoke to them. It was built around what I'd already shown I could do, not pulled from a job spec sitting in a drawer.

I start there as an AI Engineer Apprentice in August, and I'm spending the weeks beforehand the same way I spent the back half of my placement: learning things nobody's assigned me yet, currently Python, the Claude API, n8n, and Microsoft Copilot Studio.

If there's one thing I'd want a placement student to take from this, it's that the line from where you start to where you end up is rarely straight. Mine ran through Marketing Operations, campaign reports, and a traffic problem nobody asked me to fix. But the work you do sideways, outside the brief, often ends up mattering more than the brief itself.

So, just remember your career will branch out in directions you often didn't know existed, however this is how you learn what you like and so make sure you take every opportunity! Good Luck!