• Design the approach for integrating LLMs into a product workflow, including model selection and orchestration.
• Build and iterate prompting strategies, tool use, and structured outputs.
• Implement evaluation and quality measurement, including test sets, rubrics, and regression checks.
• Create an improvement loop using real user feedback and outcomes.
• Add guardrails for safety, reliability, and consistency.
• Optimise for cost, latency, and throughput using routing, batching, caching, and fallbacks.
• Work closely with the engineering lead to integrate the AI layer cleanly into the platform.
• Document configuration, prompts, evaluation methodology, and operating guidelines for handover.

• Proven experience deploying LLM features into real products, not just prototypes.
• Strong Python and practical ML engineering skills.
• Experience with several of the following:
  • RAG and embedding
  • Vector databases
  • Fine-tuning or adaptation approaches
  • Evaluation frameworks and measurement
  • Prompt injection considerations and output constraints
• Strong product mindset and comfort balancing quality, speed, and cost.

• Experience with workflow tools, internal platforms, or approval and review loops.
• Familiarity with cloud deployment patterns for AI services.

What you need to do now

If you're interested in this role, click 'apply now' to forward an up-to-date copy of your CV, or call us now.

If this job isn't quite right for you, but you are looking for a new position, please contact us for a confidential discussion about your career.