My client is looking for a high‑ownership AI Engineer to help design, build, and scale the AI systems that power our platform. You’ll work directly with our Head of AI, shaping the architecture, capabilities, and direction of the entire AI stack from retrieval to agents to model serving.

About the Role

You’ll be a core contributor to the intelligence layer of our product, working across:

• RAG pipelines: retrieval strategies, chunking, reranking, and evaluation
• AI agents: extending capabilities using MCP and our internal agent framework
• Model serving: infrastructure for open‑source and self‑hosted LLM deployment
• Developer tooling: experiment tracking, model evaluation, and broader MLOps
• Unified API layer: bringing all AI endpoints under a single interface
• Customer enablement: understanding real‑world usage and improving the system based on feedback
• Internal & external agents: building practical solutions for real customer problems

You’ll Thrive Here If…

• You understand LLMs, retrieval systems, embeddings, and agents under the hood, not just at the library level.
• You’ve shipped AI systems to production, not just experimented in notebooks.
• You’re comfortable designing backend services, APIs, and system architectures.
• You operate with high agency you take ambiguous problems, shape them, and execute.
• You care deeply about whether customers actually use and benefit from what you build.
• You prefer solving real problems over perfecting theoretical ones.

What You Bring

• Strong Python engineering skills (Go experience is a bonus)
• Experience building RAG systems, agents, or LLM‑based applications in production
• Knowledge of LLM fundamentals — transformers, embeddings, retrieval, context engineering, evaluation
• Hands‑on model‑serving experience (Hugging Face, vLLM, open‑source models)
• Familiarity with vector databases (pgvector, Milvus, Chroma)
• Interest or experience in MLOps (CI/CD, experiment tracking, evaluation)
• Comfort working with containerised environments and basic Kubernetes workflows

Nice to Have

• ML/NLP/IR publications (NeurIPS, ICLR, EMNLP, ACL, etc.)
• Experience fine‑tuning encoders
• Synthetic data generation for evaluation
• Contributions to open source
• Model compression experience (GPTQ, AWQ, etc.)
• Familiarity with multimodal models
• Deeper Kubernetes experience

✉ Email: [email protected]

☎ Contact Number: +44(0)1915949744 (1587) / +358 753 266586