What You'll Do
You'll design, build, and operationalise AI across the business:
Define and implement an AI governance framework (approval, testing, monitoring, lifecycle management)
Integrate Azure OpenAI / OpenAI APIs with proper guardrails (PII handling, audit logging, content filtering)
Build and scale AI pipelines for claims processing (OCR, classification, entity extraction, validation)
Establish prompt engineering standards and reusable prompt libraries
Create evaluation frameworks to measure accuracy, latency, and cost
Implement responsible AI practices aligned to regulatory requirements
Optimise AI cost efficiency (token usage, caching, model selection)
Support and enhance Contact Centre Copilot solutions
Roll out and manage GitHub Copilot across engineering teams
Monitor model performance in production and proactively manage drift
Engineering Contribution (Non-Negotiable)
Write and maintain production-grade Python code for ML pipelines
Contribute to platform engineering and integration efforts
Collaborate on API development and system integration
Participate in code reviews, testing, and engineering standards
Support modernisation of legacy systems where AI integrates into core workflows
What You Bring
5+ years in software engineering, with 2+ years focused on AI/ML
Proven experience deploying and maintaining AI/ML in production
Hands-on experience with Azure OpenAI or OpenAI APIs
Strong Python skills for building and maintaining ML pipelines
Experience with document AI (OCR, NER, classification, form recognition)
Understanding of LLM patterns (RAG, fine-tuning, embeddings, prompt engineering)
Familiarity with vector databases and semantic search
Strong engineering discipline: testing, monitoring, documentation, code quality
Education
Degree or Diploma in Computer Science, Data Science, Artificial Intelligence, Software Engineering, or a related technical field
Advantageous:
Certifications in AI, machine learning, or cloud platforms (Azure preferred)
Equivalent practical experience building real-world AI systems in production will be considered.