Principal Machine Learning Engineer

If you're a data scientist who actually enjoys getting things into production - not just building models - this is worth a look., In return, you get exposure to genuinely interesting problems, end-to-end ownership, and a team that values delivery over theory.

• ~6-8+ years in data science / machine learning
• Strong grounding in statistics, experimentation, and evaluation methods
• Experience deploying ML systems into production environments
• Solid engineering capability (typically Python-based stacks)
• Exposure to cloud platforms (AWS or Azure)

You'll stand out if you've worked in more complex or regulated environments - particularly across pharma, biopharma, medtech, or manufacturing - and understand what "production-ready" really means in those settings.

Additional things that tend to separate strong candidates:

• Experience working with ML in regulated environments (e.g. GAMP frameworks)
• Familiarity with regulatory standards like ISO 27001, ISO 13485, or GxP (GDP / GMP)
• Practical experience applying LLMs across real business use cases (not just experimentation)
• Exposure to manufacturing or biopharma data science problems

Where people tend to do well here:

You're pragmatic. You care more about impact than elegance. You're comfortable owning a problem properly, not just your slice of it. And you don't mind getting pulled into conversations with stakeholders when needed.

Immersum are working with a fast-growing AI consultancy delivering intelligent systems into complex, regulated environments. They're hiring Principal engineers who can own problems end-to-end: from shaping the approach, through modelling, to deploying something that genuinely works in the real world. This isn't a research role. It's applied, hands-on, and client-facing. You'll be working directly with stakeholders, building and deploying ML systems using Python and modern cloud environments (AWS / Azure), and helping organisations understand whether what they've built is actually performing - and why. The sweet spot is someone who's data science-led but engineering-capable. You'll spend your time applying statistical modelling and classical machine learning techniques, designing experiments that tie back to commercial outcomes, and ensuring what gets built can actually scale in production systems. There's an expectation you can take something from idea * prototype * production without handing it off. You'll also be in front of clients. Not in a salesy way - more in a "this is what the data says, here's what's working, here's what isn't" kind of way. Being able to explain performance, handle challenge, and build trust matters. There is also increasing exposure to AI and LLM-driven use cases, but this isn't an LLM-only role - strong fundamentals in ML and statistics are far more important than prompt engineering alone.