Applied AI Engineer (LLM Systems & Integration)

We are looking for an Applied AI Engineer (LLM Systems & Integration) to build, integrate, evaluate, and optimize secure AI applications within the Bank’s controlled technology environment. You will connect approved LLMs and AI services with backend systems, enterprise data sources, and operational workflows. Key responsibilities include developing retrieval and context-assembly pipelines, production-grade APIs, and AI evaluation frameworks. Working with architecture, infrastructure, data, security, and business teams, you will ensure AI solutions are reliable, secure, observable, and aligned with enterprise standards. You will also help establish reusable engineering and operational practices that support the Bank’s progression from AI-assisted solutions to controlled agentic capabilities. The role will translate AI capability and solution architecture into production-ready implementations and contribute to establishing reusable AI engineering, evaluation, integration, and operational practices that support the Bank’s progressive development from AI-assisted solutions toward controlled agentic capabilities.

Design, develop, and optimize production-grade LLM applications and RAG pipelines using vector, keyword, and hybrid retrieval techniques. Develop controlled agentic AI workflows with tool integration, defined autonomy levels, Human-in-the-Loop (HITL) approval gates, authorization controls, and exception handling. Build secure and governed AI data-access and context-engineering pipelines, ensuring appropriate data permissions, ownership, metadata usage, and efficient context selection. Integrate AI services with enterprise platforms and API gateways, ensuring schema validation and clear separation between probabilistic AI outputs and deterministic business logic. Implement reliable integration patterns, including retries, timeouts, error handling, idempotency, and health monitoring. Establish AI evaluation and quality frameworks to measure accuracy, relevance, hallucinations, structured-output reliability, and regression performance. Benchmark and optimize model and inference performance, including latency, TTFT, throughput, concurrency, and context-length requirements. Implement comprehensive AI observability, traceability, and version control across models, prompts, context, tool calls, and execution logs. Ensure AI solutions comply with information security, privacy, data-retention, and operational-control requirements, while continuously monitoring quality after deployment.

Requirements 3+ years of professional backend software engineering experience, preferably building data-intensive applications or AI integrations. Strong proficiency in Python or C#/.NET, with solid knowledge of RESTful APIs, software design patterns, and relational databases (SQL). Hands-on experience with LLM application development, including prompt engineering, context management, structured outputs, RAG, and model evaluation. Practical understanding of Human-in-the-Loop (HITL), tool/function calling, and controlled AI autonomy patterns. Strong knowledge of software testing, CI/CD, version control, logging, and production application support in security-conscious enterprise environments. Ability to translate complex business and process requirements into production-ready AI solutions. Preferred Qualifications Experience with advanced RAG and enterprise search, including vector databases, embedding models, reranking, and search engines. Experience deploying or benchmarking LLMs and inference runtimes in private, on-premises, or secure enterprise environments. Understanding of data warehousing, dimensional modeling, semantic layers, and data governance. Experience in banking, financial services, or other highly regulated industries. Experience with LLMOps, AI evaluation and regression-testing frameworks, observability, tracing, and production AI monitoring.