LLM Data Scientist

Holland and Barrett
London
1 day ago
Create job alert

If you're someone who loves the intersection of LLM systems, experimentation, and realworld product impact, this is the kind of role where your work won't stay hidden. You'll see it live, in customerfacing journeys, SEO performance, personalisation, and in the quality of the content we publish.


And to be clear: this is not a role for someone who has only ever engineered prompts in isolation. You'll definitely write prompts - but the value here comes from understanding how LLMs behave in real systems, with real data, and knowing how to evaluate, measure, optimise and deploy them at scale. If you have the data science mindset and the LLM engineering skillset, this is where you can genuinely shape things.


What you'll actually be doing?



  • Designing, testing, and optimising prompts and promptchains for production LLM applications
  • Building evaluation frameworks to measure quality, safety, relevance, latency and cost
  • Working with RAG architectures, embeddings and vector databases to improve retrieval and contextual accuracy
  • Owning and developing our SEO contentoptimisation tool - improving product page copy for both relevance and discoverability
  • Helping power personalised recommendations and messaging in our Personalisation squad
  • Enhancing our Wellness Hub (our library of health articles and expert content) by generating, rewriting and improving pieces in safe, evidencebased ways
  • Partnering with Data Engineering on context management, memory strategies, tool/functioncalling, observability and logging
  • Monitoring and mitigating model drift, prompt degradation and system regressions
  • Acting as a goto LLM expert across the business, helping teams understand what's possible and what's not
  • Staying close to the evolving LLM landscape and translating new ideas into useful, practical improvements

How we'll see your impact



  • Customers engage more with product pages because the content is more helpful and tailored
  • SEO performance improves thanks to your optimisation tools
  • We publish higherquality health content that's accurate, engaging and aligned with our brand
  • Customers receive more relevant recommendations at the right moment
  • Internal teams work faster thanks to stronger, more reliable LLMpowered tools
  • Clear evaluation metrics guide smarter, datadriven AI adoption
  • LLM capability across H&B grows because you've helped demystify the technology
The Person

What you'll need



  • Handson experience developing LLM applications beyond just prompt writing
  • Solid data science foundations: experimentation, evaluation, statistical thinking, working with realworld datasets
  • Strong understanding of: prompt optimisation, RAG, embeddings, vector stores, context management and LLM evaluation
  • Very strong Python skills - this is a genuinely handson role
  • Ability to reason about tradeoffs between quality, latency and cost
  • Experience with Git, version control and CI/CD
  • Confidence translating complex LLM concepts for nontechnical colleagues
  • Curiosity, pragmatism, and comfort shaping a firstofitskind role
Benefits





 Technology Incentive Scheme - we offer different bonus schemes for all grades in Technology, starting at 10%.


 Learning and Development opportunity with Holland & Barrett is a great base for career development long term.


Career progression


 Pension company contribution 


Your wellbeing is paramount so you can get away and take 28 or 33 Days Holiday per year. 


 Refer and Earn Scheme - as we're growing you can earn money by referring people to join us from your network.


 Epic Extras gives you access to exclusive benefits, free advice and savings from a range of retailers and providers.


 Stay healthy with Discounted Products - from day one you'll get a 25% discount (on top of other promotions) when you shop at H&B on anything that you buy. 


 We all need a little help sometimes, so we offer Free 24/7 Confidential Advice & Colleague Welfare.


 Mental Health First Aiders - we have lots of qualified Mental Health First Aiders because its all about your health & wellbeing.


We have colleague Reward and Recognition Schemes, so your hard work and loyalty won't go unnoticed.


And many more!



Related Jobs

View all jobs

LLM Data Scientist/Engineer (Security-Cleared, Onsite)

Remote NLP & LLM Data Scientist for AI Analytics Platform

Lead Data Scientist

FinTech Data Scientist: ML & LLM Analytics

Principal Data Scientist - Agentic AI & LLM Engineering

Lead Data Scientist – Generative AI & LLM Expert

Subscribe to Future Tech Insights for the latest jobs & insights, direct to your inbox.

By subscribing, you agree to our privacy policy and terms of service.

Industry Insights

Discover insightful articles, industry insights, expert tips, and curated resources.

How Many AI Tools Do You Need to Know to Get an AI Job?

If you are job hunting in AI right now it can feel like you are drowning in tools. Every week there is a new framework, a new “must-learn” platform or a new productivity app that everyone on LinkedIn seems to be using. The result is predictable: job seekers panic-learn a long list of tools without actually getting better at delivering outcomes. Here is the truth most hiring managers will quietly agree with. They do not hire you because you know 27 tools. They hire you because you can solve a problem, communicate trade-offs, ship something reliable and improve it with feedback. Tools matter, but only in service of outcomes. So how many AI tools do you actually need to know? For most AI job seekers: fewer than you think. You need a tight core toolkit plus a role-specific layer. Everything else is optional. This guide breaks it down clearly, gives you a simple framework to choose what to learn and shows you how to present your toolset on your CV, portfolio and interviews.

What Hiring Managers Look for First in AI Job Applications (UK Guide)

Hiring managers do not start by reading your CV line-by-line. They scan for signals. In AI roles especially, they are looking for proof that you can ship, learn fast, communicate clearly & work safely with data and systems. The best applications make those signals obvious in the first 10–20 seconds. This guide breaks down what hiring managers typically look for first in AI applications in the UK market, how to present it on your CV, LinkedIn & portfolio, and the most common reasons strong candidates get overlooked. Use it as a checklist to tighten your application before you click apply.

The Skills Gap in AI Jobs: What Universities Aren’t Teaching

Artificial intelligence is no longer a future concept. It is already reshaping how businesses operate, how decisions are made, and how entire industries compete. From finance and healthcare to retail, manufacturing, defence, and climate science, AI is embedded in critical systems across the UK economy. Yet despite unprecedented demand for AI talent, employers continue to report severe recruitment challenges. Vacancies remain open for months. Salaries rise year on year. Candidates with impressive academic credentials often fail technical interviews. At the heart of this disconnect lies a growing and uncomfortable truth: Universities are not fully preparing graduates for real-world AI jobs. This article explores the AI skills gap in depth—what is missing from many university programmes, why the gap persists, what employers actually want, and how jobseekers can bridge the divide to build a successful career in artificial intelligence.