Machine Learning Engineer / Data Scientist – LLM Agents

Springer Nature; London; On-site; GBP 60,000 - 80,000; Full time

Job summary

A global leader in research and education is looking for a Machine Learning Engineer in London to work on large language models, build multi-step agents, and enhance AI products. Ideal candidates will have a strong background in data science, proven experience with LLM fine-tuning, and strong Python skills. This role requires a keen problem solver who thrives in fast-paced research environments.

Qualifications

• Strong data science expertise and LLM experience required.
• Proven track record with large language models in production environments.
• Hands-on experience with automated testing and best practices.

Responsibilities

• Design and build LLM agents with LangGraph.
• Create and deploy ML pipelines for model training and fine-tuning.
• Collaborate with scientists to turn research ideas into products.

Skills

Python programming

Experience with PyTorch

Knowledge of HuggingFace Transformers

Container management

Reinforcement learning

Education

BSc, MSc or PhD in Computer Science

Tools

LangGraph

GCP

SQL

Job description

Role Description

We are looking for a machine learning engineer with strong data science expertise to join the team working on large language models for life and natural science problems. Work involves building agentic workflows where LLMs reason, plan and act, as well as developing pipelines to train and fine-tune models. LangGraph is our main framework for agent development; knowledge of other agent stacks is a plus

Key Responsibilities

• Design and build multi-step LLM agents with LangGraph and similar frameworks

• Create data and ML pipelines for continual pre-training, supervised fine-tuning and RL alignment

• Deploy models and retrieval services on containerised infrastructure with reliable CI/CD

• Monitor and improve agent performance with Weights & Biases and internal dashboards

• Collaborate with scientists and engineers to turn research ideas into working products

Required Qualifications

• BSc, MSc or PhD in Computer Science, Data Science or a related field

• Strong Python skills with PyTorch, HuggingFace Transformers and Datasets

• Proven track record fine-tuning and serving large language models in real-world settings

• Hands-on experience building pipelines with reinforcement-learning algorithms such as PPO and GRPO

• Competence with containers, automated testing and software-engineering best practice

Useful Skills

• Basic experience with GCP and infrastructure-as-code workflows

• Hands-on experience using vector, graph and relational databases, plus SQL and data modelling

• Experience with multimodal models and emerging agent protocols such as MCP and A2A

• Ability to implement model safety and guard-rail measures

Personal Attributes

• Team player with clear communication

• Analytical and detail-oriented problem solver

• Curious and quick to learn new methods

• Comfortable in a fast-moving research environment

• Committed to delivering maintainable, reliable software

#LI-SS2