Senior LLM & Multi-Agent Systems Engineer

Contract Type: Outside IR35

Day Rate: -375 - -400 per day

Senior LLM & Multi-Agent Systems Engineer

Location: Flexible (UK or Europe preferred)

Type: Permanent / Contract

Industry: AI, NLP, and Applied Machine Learning

We are seeking a highly experienced AI engineer with deep expertise in large language models (LLMs), multi-agent systems, and emerging prompt and context engineering practices. The ideal candidate will be fluent in the latest industry developments, capable of recalling and applying detailed model specifications, and confident in designing, optimising, and deploying advanced LLM-based architectures for production use.

You will work on building scalable, intelligent systems that leverage agentic workflows, advanced reasoning techniques, and multi-agent communication architectures, integrating them into robust, performant, and explainable AI pipelines.

Key Responsibilities

• Architect and implement LLM-based systems with advanced reasoning capabilities, including context engineering, role-based prompting, and multi-agent orchestration.
• Select and integrate appropriate frameworks and tools (e.g. LangChain, LangGraph, AutoGen, CrewAI, Agno, Smallagons, OpenHands) for specific use cases.
• Design, optimise, and evaluate tokenisation, embedding, and vector store strategies - including trade-offs between embedding dimensions, indexing performance, and retrieval costs.
• Implement advanced multi-agent architectures for collaborative AI workflows, selecting the appropriate communication protocols and frameworks.
• Integrate reasoning methods (e.g. probabilistic reasoning, causal reasoning, knowledge graph integration) into LLM pipelines to improve reliability and explainability.
• Keep pace with the latest developments in LLM research, frameworks, and best practices, proactively applying them to ongoing projects.
• Work closely with product and engineering teams to align AI system design with business objectives.
• Conduct rigorous evaluation of LLM performance, context window trade-offs, and hallucination mitigation techniques.

Essential Skills & Experience

• Proven track record designing and deploying production-grade LLM solutions.
• Expert understanding of context engineering and prompt design patterns for multi-modal, multi-agent environments.
• Strong knowledge of multi-agent frameworks (CrewAI, AutoGen, LangGraph, Agno, OpenHands, etc.) and when to use them.
• In-depth knowledge of tokenisation strategies (e.g. BPE, SentencePiece, byte tokenisation), embedding models (e.g. OpenAI Ada, text-embedding-3-large), and their performance implications.
• Familiarity with knowledge graphs (e.g. Neo4j, LangDB, in-memory graphs) and graph neural networks for reasoning.
• Strong grounding in probabilistic models, causal inference, and AI reasoning approaches.
• Hands-on experience with cloud platforms (AWS, GCP, Azure) and MLOps workflows.
• Solid understanding of vector databases (e.g. Pinecone, Weaviate, Milvus) and retrieval-augmented generation (RAG) techniques.
• Comfort with rapid technical recall for industry benchmarks, model parameters, and framework capabilities.