Senior NodeJS AI Engineer

We are looking for a quick-learning engineer with proven production AI adoption to implement advanced RAG, agent-based capabilities, robust API and MCP integrations, and rigorous evaluation/testing. The focus is AI architecture and reliability first, with solid backend, CI/CD, observability and SAP HANA via TypeORM. Responsibilities Design and implement agentic capabilities such as tool/function calling, planning, memory/state, safety guardrails and failure recovery for production stability Build advanced RAG pipelines on OpenSearch including hybrid retrieval, index/chunking strategies, embeddings/rerankers, query rewriting and retrieval evaluation Development and maintenance of API and MCP integrations, implementation of MCP servers/clients, capability manifests, tool adapters and clean versioned REST/WebSocket APIs Creation and iteration of prompt strategies covering instruction/few-shot design, tool-aware prompting, structured outputs/JSON and prompt governance/guardrails Establish AI quality evaluation and testing with RAGAS, DeepEval (or similar), regression suites, red teaming and offline/online metrics Implement data access with SAP HANA using TypeORM including schema modeling, performant queries, migrations and secure data handling Ensure observability and performance through service instrumentation, latency/cost/quality tracking with Dynatrace, SLIs/SLOs definition and optimization of LLM workloads Build and maintain CI/CD for Kyma/Kubernetes including automated tests, security scanning, container build/versioning, safe rollbacks and environment configuration Collaboration with product and architecture, documentation of designs and participation in reviews Requirements 5+ years of experience as a NodeJS engineer with proven production AI adoption Expertise in agentic capabilities including tool/function calling, planning and memory/state Proficiency in advanced RAG pipelines on OpenSearch with hybrid retrieval, embeddings/rerankers and query rewriting Skills in API and MCP integrations including MCP servers/clients, tool adapters and REST/WebSocket APIs Competency in prompt strategies covering instruction/few-shot design, structured outputs/JSON and prompt governance Background in AI quality evaluation with RAGAS, DeepEval and regression suites Knowledge of SAP HANA, TypeORM and secure data handling Familiarity with observability and performance using Dynatrace, SLIs/SLOs and LLM workload optimization Understanding of CI/CD for Kyma/Kubernetes including automated tests, security scanning and safe rollbacks Nice to have Knowledge of SAP Joule