LLM Data Scientist (up to £120k + Equity)

Role: LLM Data Scientist (Forward-Deployed)

Location: London (On-site; Liverpool Street)

Employment Type: Full-time and Permanent

Remuneration: £75k - £120k Base Salary + Discretionary Bonus + Equity

Zettafleet is an end-to-end platform for businesses and organisations to train their own LLM on their proprietary data. We can use non-conventional AI hardware and automatically source and combine GPUs (and other types of AI accelerators) from multiple cloud providers, enabling users to optimise for cost, duration or geographic location of the training.

The founding team consists of Oxford and Cambridge graduates and former engineers at Google, Meta, Microsoft and Amazon. We are backed by prominent investors from the US and the UK, including institutional VC funds and C-level executives of global technology companies.

We are looking for a high-impact Forward-Deployed LLM Data Scientist to bridge the gap between enterprise data environments and advanced AI models. In this high-visibility, customer-facing technical role, you will embed with client teams to turn their raw data into clean Parquet files for training of domain-adapted LLMs and embedding models. You will act as both a data scientist and consultant, advising clients on the complete data lifecycle - from dataset creation and synthetic data generation to identifying the right data retrieval and LLM training strategies.

In this role, you will:

• Partner directly with customer engineering leadership and product teams to understand their business roadmap, technical requirements and existing data infrastructure.
• Build data pipelines to transform raw enterprise data into training-ready formats.
• Design and implement data cleaning, filtering, deduplication and synthetic data generation strategies tailored for Continued Pre-Training (CPT), Supervised Fine-Tuning (SFT), Direct Preference Optimisation (DPO) and Reinforcement Learning (RL) workflows.
• Continuously evaluate model input and output quality, iteratively refining dataset preparation and training strategies to optimise performance.
• Work with Zettafleet's in-house engineering, research and data teams to solve non-trivial data challenges and edge cases.
• Capture insights, friction points and feature requests to directly influence Zettafleet's core product roadmap.
• Occasionally travel to the customer's place of work, whether it's in the UK or overseas.

What we are looking for:

• LLMs: Practical experience with LLM training/fine-tuning and model evaluation. Good applied understanding of modern LLM concepts (e.g., synthetic data generation, data tokenisation, pre-training, post-training, vector databases and RAG).
• Data Science: 1+ years of professional experience in Data Science or Machine Learning using Python. Overall, a minimum of 2 years of work experience which involved programming.
• Problem solving: Strong analytical problem-solving skills and attention to detail.
• Strong ownership mindset: Ability to take full responsibility for outcomes, not just tasks, and great care about the quality of deliverables.
• Collaboration and communication: Genuinely exceptional verbal and written communication skills, capable of translating complex Data Science concepts into actionable strategies for executive and technical audiences alike.

We would like to acknowledge that almost no candidate checks every box - and that is perfectly fine. If you are passionate about solving complex challenges and open to learning new technologies, we would love to hear from you.

Nice to have:

• Proven track record in a customer-facing technical role (e.g., forward-deployed engineering, solutions architecture or technical consulting) with strong stakeholder management abilities.
• Reasonable proficiency in one additional area (e.g., frontend, backend, distributed systems, applied research or UI/UX/visual design).
• Prior experience in a fast-paced AI startup or high-growth technology company.
• Contributions to and experience with open-source projects.

How we work:

• We are a small, early-stage team with big ambitions. We move quickly, care deeply about what we build, and take pride in doing things well.
• This is a high-ownership environment where people who are proactive, curious, and motivated by impact tend to thrive.
• Work in an environment conducting cutting-edge research in AI.
• Competitive salary, equity and benefits package.
• 28 days + public holidays allowance.
• Opportunities for professional growth and progression with your career.
• Work on challenging engineering problems that have a real impact on the industry.
• Work with high-profile customers and technology partners.

#J-18808-Ljbffr