AI Engineer (LLM / Agentic Systems)
Dear applicants, please keep in mind that applications without provided salary expectations and active LN profile will not be considered. Hope for your understanding.Location: San Francisco, CA (On-site)Employment Type: Full-TimeBenefits: 100% medical, dental, vision; MacBook Pro + peripheralsWe are hiring an AI Engineer to own the intelligence layer. This is not a demo or prototype role. This is production AI engineering.About The CompanyClient builds production AI agents that replace manual coordination work inside billion-dollar enterprises. Our agents operate at scale — processing thousands of transactions, making classification decisions, routing exceptions, and learning from human feedback.We deploy intelligent agents directly into enterprise systems such as:SalesforceNetSuiteServiceNowWorkdayYou will:Design and ship agentic systems used in real enterprise workflowsBuild evaluation and reliability systemsHandle hallucinations, edge cases, cost constraintsOptimize multi-agent orchestration in productionYou should already have built LLM-powered systems that operate beyond the playground stage.What You’ll OwnAgent architecture designRetrieval systems (RAG, context management)Tool calling and multi-step reasoningMulti-agent orchestrationPrompt engineering and reliability optimizationEvaluation and quality infrastructureCost-performance tradeoff optimizationException routing and human-in-the-loop feedback loopsMust-Have Requirements3+ years software engineering experience2+ years building production LLM or AI systemsHands-on experience with agentic workflowsExperience with tool calling, retrieval, and multi-step reasoningStrong prompt and context engineering skillsExperience building evaluation frameworks for AI outputsStrong Python and backend fundamentalsExperience handling hallucinations, edge cases, and cost controlBased in San FranciscoNice to HaveExperience integrating AI into enterprise SaaS systemsExperience with vector databasesExperience designing HITL systemsExperience with scaling AI workloads