> Markdown version of [/jobs/ext/4757635-ai-migration-engineer](https://www.wearedevelopers.com/jobs/ext/4757635-ai-migration-engineer). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # AI Migration Engineer - **Company:** HSBC Group - **Location:** Sheffield, United Kingdom - **Contract:** Permanent contract - **Skills:** Application Programming Interfaces (APIs), Artificial Intelligence, Amazon Web Services, Automation of Tests, Microsoft Azure, Spreadsheets, Code Review, Continuous Integration, Python (Programming Language), Machine Learning, Retrieval-Augmented Generation, Large Language Models, Generative AI, Git, Data Lineage, Data Management, Graph RAG, Code Restructuring - **Published:** November 17, 2026 - **Apply:** https://dejobs.org/x/x/710A8D07DA0B4683AC473C358F5A5EBA/job/ ## About the Role * Bring substantial hands-on experience in software data or ML engineering including proven delivery of LLM-based applications into production * Demonstrate strong Python engineering with experience across APIs containers CI/CD and cloud platforms such as Azure GCP or AWS * Build RAG pipelines and retrieval approaches including vector and hybrid search and apply agent orchestration frameworks in production settings * Create evaluation and monitoring approaches that cover accuracy faithfulness robustness regression drift quality and guardrails * Show proven depth in at least one of these areas with working knowledge across the others - skills and monitoring or orchestration and MCP or tools refactoring or deployment and GUI * Apply strong engineering practice including automated testing code review Git and reliable CI/CD ways of working * Handle data confidently across structured and unstructured sources including documents PDFs and emails with strong SQL capability * Deliver within a regulated environment with experience evidencing controls for audit and regulators and working within model risk data protection and change governance ## Description * Map Risk & Compliance processes end to end with process owners including inputs hand-offs data lineage controls and exceptions * Simplify processes before automation by removing redundant steps duplicate controls manual reconciliations and rework loops * Re-engineer target processes around agentic patterns with explicit human-in-the-loop decisions and control points * Build and integrate agentic workflows using LLMs tool calling and MCP connecting to Risk & Compliance systems data platforms and document stores * Implement retrieval patterns including RAG GraphRAG and hybrid search grounded in Risk & Compliance policy regulatory and risk data * Replace or wrap legacy components such as spreadsheets macros scripts and EUCs with maintainable version-controlled code * Build evaluation harnesses for accuracy faithfulness robustness and regression using golden datasets agreed with the business * Embed controls and deploy via LLMOps pipelines with monitoring for drift quality cost and latency plus parallel runs cutover and rollback plans