Senior / Staff AI Software Engineer

7+ years of backend development experience in production environments, specifically strong Python skills, including async programming and type hints. Experience building and monitoring production-quality ML and AI systems. Hands-on expertise with RAG frameworks and agentic workflows. Solid understanding of PostgreSQL database design and optimization. Familiarity with Docker and containerization. Strong testing practices using pytest. Experience with microservice architectures is preferred. Experience with GraphQL APIs. Experience with event-driven systems and message queues. Experience with major cloud providers (e.g., AWS). Shipped production LLM systems: ideally RAG architectures, agent/tool use, and prompt/system design, with LangChain (tracing/evals via LangSmith), embeddings, and vector databases (Databricks Vector Search preferred). Deep expertise in retrieval quality, including chunking, metadata, hybrid search, reranking, and grounding.

Design and develop robust, scalable, event-driven services using Python, FastAPI, Apache Kafka and GraphQL. Build fundamental LLM agents and integrate them into our product. Work with DevOps on deployments, monitoring, and reliability improvements. Maintain and optimize PostgreSQL databases and data models. Collaborate across product and engineering teams to define requirements and architect features. Drive engineering best practices through code reviews and mentorship. Engage with current and prospective clients to drive understanding of the Caregentic AI architecture and capabilities. Design, build, and operate LLM services, including RAG systems (LangChain), agentic workflows, and evaluation pipelines (LangSmith, deepeval, A/B testing). Own vector search & embeddings pipelines from schema and metadata design to model benchmarking, cost/latency optimization, and Databricks Vector Search integration. Lead conversational AI development enhancing NLU policies, safety guardrails, and custom action servers, plus integrating assistants with microservices.