AI & Data Engineer in Manchester

You will operate at the intersection of Data Engineering, Data Science, and modern AI/ML systems, taking ownership of initiatives that directly shape product and business outcomes. We need someone with genuine breadth — equally comfortable designing scalable data pipelines as they are building agentic AI architectures — and with the curiosity to keep pace with a space that is moving faster than almost any other in software engineering. You will bring deep technical expertise across the full AI/ML stack alongside the leadership qualities to mentor colleagues, challenge assumptions, and drive a culture of engineering excellence. Crucially, you will not just set direction — you will get your hands dirty and build it too.

What You Will Be Doing

• Lead the refactoring of legacy infrastructure into highly scalable, secure, and multi-tenant data pipelines that power our Security Portfolio Intelligence platform.
• Own data quality, governance, and security end-to-end: establishing robust validation frameworks, automated alerting, and compliance-ready data modeling for highly regulated enterprise clients.
• Champion pragmatic AI: ruthlessly identify where LLMs can be replaced by leaner, more cost-effective classical ML models, ensuring optimal performance and cost-efficiency.
• Evolve our data architecture: champion the appropriate pattern for the job, whether that's a GraphDB, vector stores, and more standard SQL/NoSQL structures, all whilst ensuring scalability and long-term maintainability.
• Establish and promote good data modelling practices across the organisation — schema design, query optimisation, and a sensible approach to data governance.
• Work across a range of storage paradigms: SQL (PostgreSQL, MySQL), vector databases (pgvector and equivalents), and graph databases (Neo4j or similar).
• Design and ship agentic AI pipelines and multi-agent reasoning systems that solve real business problems — content review, classification, enrichment, and beyond.
• Lead the evaluation and adoption of emerging AI/ML tooling: Vertex AI, Google ADK, AWS SageMaker, Azure ML, and next-generation LLM frameworks.
• Establish LLMOps practices: formal evaluation pipelines, regression testing, and quality baselines so we always know whether our AI systems are improving or declining.
• Identify where Large Language Models can be replaced by leaner, more cost-effective traditional ML models — and deliver those replacements.
• Build NLP-powered systems, including classifiers, semantic search, and potentially fine-tuned or custom-trained models where the use case justifies it.
• Drive the auto-generation of marketing content and other AI-powered product features, working closely with Product to turn ideas into production systems.
• Bring your own ideas to the table. If you see an opportunity we have not spotted, we want to hear it — and we will give you the space and support to explore it.

Leadership & Cross-Cutting Responsibilities

• Mentor and collaborate with engineers across the team, raising the collective bar for AI/ML quality, reproducibility, and best practice.
• Run tech-sharing sessions; keep the team current on fast-moving developments in the AI/data space.
• Contribute to hiring: interview, assess, and help build the team you want to work in.

Requirements (Must-Have)

• 7+ years of professional experience across Data Engineering, Data Science, or Machine Learning roles — with meaningful exposure to both the data and AI/ML sides of that spectrum.
• Hands-on experience designing and shipping agentic AI systems and multi-agent architectures (LangChain, LangGraph, AutoGen, Google ADK, or similar frameworks).
• Strong working knowledge of Large Language Models: prompt engineering, evaluation, and responsible deployment in production.
• Experience with Cloud ML platforms — at least one of Vertex AI (GCP), SageMaker (AWS), or Azure Machine Learning.
• Expertise in Python for data processing, model training, and API development.
• Solid understanding of classical ML and NLP: ability to identify when a simpler model outperforms an LLM in production and to deliver that alternative.
• Relational databases: PostgreSQL, MySQL, or equivalent — schema design, query optimisation, and data modelling.
• Vector databases: practical production experience with pgvector, Pinecone, Weaviate, or similar for semantic search and RAG pipelines.
• Graph databases: Neo4j or equivalent; experience modelling domain knowledge as a graph.
• Demonstrated technical leadership — not necessarily formal line management, but clear ownership of complex technical workstreams and influence over engineering decisions.
• Strong communication skills; comfortable translating technical concepts for non-technical stakeholders.

Nice-to-Have (Bonus Points)

• Cybersecurity domain knowledge, specifically an understanding of how enterprise security tools map to frameworks like NIST and MITRE.
• GraphQL API design and implementation.
• MLOps / LLMOps tooling: MLflow, Weights & Biases, Evidently AI, or similar for experiment tracking and model monitoring.
• Experience training or fine-tuning your own models (transformer-based or otherwise) from scratch or from pre-trained checkpoints.
• NLP specialism: named entity recognition, text classification, semantic similarity, topic modelling, or conversational AI.
• Data orchestration tools: Airflow, Prefect, Dagster.
• Experience working with Knowledge Graphs or ontologies in a production environment.
• Published work, open-source contributions, or a track record of writing or speaking about AI/ML topics.

Engineering Manchester, UK (UK-Based Remote Considered) Full-time