About this role

Build and own production AI systems end-to-end, including LLM-powered applications, intelligent agents, conversational AI, document processing, decision-support tools, and automation workflows, from architecture and prototyping through deployment, monitoring, and continuous improvement.
Design agent, RAG, and context-engineering systems, including retrieval, embeddings, reranking, grounding, memory, compaction, caching, token management, tool/function calling, agent harnesses, retries, state management, and failure recovery.
Build robust AI evaluation and quality frameworks to measure accuracy, relevance, hallucination, safety, retrieval quality, latency, cost, and task completion, using regression testing, human evaluation, LLM-as-judge, and production feedback to continuously improve system performance.
Evaluate, train, deploy, and optimise AI models, selecting between frontier and open-weight models and applying techniques such as SFT, LoRA/QLoRA, preference tuning, quantisation, and self-hosted inference where appropriate, while optimising for capability, security, privacy, latency, and cost.
Engineer reliable and secure production AI infrastructure, developing Python/FastAPI services, APIs, data and inference pipelines, and observability while collaborating with Product, Software, and Platform Engineering teams and maintaining appropriate technical documentation, security controls, and responsible AI practices.

3+ years of relevant experience in AI Engineering, Machine Learning Engineering, Applied AI,
Data Science, or related roles, with demonstrable experience designing,
shipping, and maintaining LLM-powered systems in production.
Strong proficiency in Python,
FastAPI, APIs, SQL/databases, Git, Docker, and Linux, with solid
software engineering fundamentals including asynchronous systems, retries,
state management, testing, observability, debugging, scalability, latency,
and cost optimisation.
Hands-on experience with LLMs,
Generative AI, RAG, agents, and context engineering, including prompt
engineering, embeddings, vector/hybrid search, retrieval, reranking,
tool/function calling, memory, context compaction, caching, grounding, and
token management.
Demonstrable experience building AI
evaluation and model engineering pipelines, including regression
testing, LLM-as-judge/human evaluation, hallucination mitigation, model
benchmarking, and working with open-weight models through techniques such
as SFT, LoRA/QLoRA, DPO/preference tuning, and production inference.
Experience with relevant AI/ML
and production technologies such as PyTorch, Hugging Face, vector
databases, commercial and self-hosted LLMs, and inference frameworks such
as vLLM/SGLang, with a strong understanding of AI security, privacy,
responsible AI, model selection, performance, and production reliability.

AI Engineer