AI Engineer

AI Engineer — Agentic Workflows & Production Systems Full-Time | Enterprise Technology Division ABOUT THE ROLEWe are seeking a skilled AI Engineer to design, build, and deploy production-grade agentic AI systems in an enterprise environment. This role sits at the intersection of software engineering, machine learning infrastructure, and applied AI — you will architect intelligent workflows, implement retrieval-augmented generation (RAG) pipelines, and deliver scalable, monitored solutions that drive business impact.Experience with risk management or financial services is a plus, but not required. What matters most is a track record of shipping AI agents and workflows to production.KEY RESPONSIBILITIES• Design and implement multi-agent and single-agent agentic workflows applying established patterns (ReAct, CoT, STAR, OODA, HTN), making principled decisions about when to decompose tasks across agents versus consolidating with multiple tools• Build, optimize, and maintain RAG pipelines — including hybrid search, retrieval re-ranking, and prompt optimization — using LangGraph with state/checkpointing (AsyncPostgresSaver)◦ Manage prompt drift and token consumption at scale◦ Implement vector database re-ranking strategies to improve retrieval relevance◦ Engineer chunking, embedding, and indexing pipelines using Vectorize and similar platforms• Develop enterprise-grade Python and Java services with full test coverage, logging, and observability via OpenTelemetry and Grafana/Tempo• Integrate AI components into CI/CD pipelines and ensure smooth, automated deployments to production• Write scalable, maintainable code following engineering best practices — including unit, integration, and regression testing• Manage context window usage strategically: optimize prompt engineering to maximize model performance while controlling costs• Implement hallucination detection and mitigation strategies within agentic systems using guardrails, LLM-as-a-judge, and the Galileo eval framework; maintain offline eval pipelines and regression suites• Build and integrate MCP (Model Context Protocol) servers using mcp/fastmcp and LangChain MCP adapters for tool integration and external data accessREQUIRED SKILLS & EXPERIENCE• Hands-on experience building and deploying agentic AI workflows to production environments• Strong Python proficiency; Java experience is a significant plus• Deep familiarity with vector databases (e.g., Pinecone, Weaviate, Cloudflare Vectorize, pgvector) and embedding strategies• Practical experience with RAG architectures, including retrieval re-ranking, hybrid search, and prompt optimization• Solid prompt engineering skills — understanding of system prompts, few-shot prompting, chain-of-thought, and context window management• Experience with CI/CD pipelines (GitHub Actions, Jenkins, or equivalent) for AI/ML systems• Knowledge of hallucination risks and mitigation techniques in LLM-based applications• Software engineering fundamentals: testing, monitoring, alerting, and documentation in enterprise settingsCommon Skills:• AMP and Claude Code• PythonAI Engineers:• LangGraph + LangGraph state/checkpointing (AsyncPostgresSaver)• MCP (mcp/fastmcp) + Langchain-mcp-adapters• Fastapi, SSE streaming, async Python• Agent patterns: ReAct, CoT, STAR, OODA, HTN• Guardrails & LLM-as-a-judge• OpenTelemetry, Grafana/Tempo• Mongo DB/ Elasticsearch clients, Kafka, TeamsBot Framework• Short-term: LangGraph checkpoint state, message-history windowing summarization• Long term: Mem0 – extraction, storage, retrieval of user facts/preferences; cross-session recall; memory consolidation• Feedback loops: capturing thumbs up/down, labeled Q/A corpus• Agent eval: Galileo eval framework, offline eval pipelines, regression suites, hallucination detection