Most AI jobs out there fall into two categories:

1. Endless prototypes that never ship, or
2. AI strategy roles where nothing gets built.

This one is neither.

You’ll be joining a company where AI isn’t a side project. It’s the direction of travel. They’ve got an established products, plenty of customers, real data, and a backlog of AI problems that need solving properly, not with another POC that gathers dust.

So they’re not looking for theorists, hype merchants, or people who’ve lived their whole career inside Jupyter notebooks. They want who's got the scars and learnt through doing (and failing), someone who knows how to get things out the door and keep them running.

What you’ll be doing
You won’t be researching the future of artificial intelligence. You’ll be building things for real humans today.

• Designing and building machine learning systems from scratch, the whole thing, not just the model.
• Working heavily with LLMs and foundation models: classification, document understanding, forecasting, the fun stuff and the fiddly stuff.
• Getting features into production, not almost there if product just signs off.
• Deploying, monitoring, evaluating, retraining, and making sure it all behaves itself in production.
• Using AWS properly: S3, SageMaker, Bedrock, Lambda, the lot.
• Helping shape the MLOps patterns other teams will follow.
• Working with product people who care about customer problems, not academic benchmarks.
• Being a grown up voice in a team that is still forming its habits and standards.

If you like being thrown ambiguous problems and figuring out what to do, you’ll enjoy this. If you need a ticket for every decision, you won’t.

Who this suits

• Ideally you're a software engineer first, ML engineer second.
• You write reliable Python. You understand systems. You know how cloud infrastructure works. And you know that a model with great metrics is useless if the rest of the pipeline falls over.
• You’ve shipped AI powered features before. Not just POCs. Not just assisted with. Actual features that customers used.
• You’re comfortable with LLMs, transformers, evaluations, guardrails, MLOps, containers, Kubernetes, the usual suspects. You don’t need to be an expert in all of them. But you shouldn’t be surprised by any of them.
• You can explain things to people who don’t care how attention mechanisms work.
• You don’t crumble when something breaks.
• And you don’t get precious when someone challenges your idea.
• You’re curious enough to keep up with the AI world, but grounded enough not to chase every shiny tool that pops up on Twitter.

What’s in it for you

• You get to help build the AI foundation of a company that is changing how it works, not just changing its job titles.
• You’ll work with senior engineers, product thinkers, and a leadership group that wants to ship things. Not in five years. Now.
• You’ll have influence. You’ll have ownership. And you’ll have the kind of problems that make you better at your job.