About the RoleThis is a mid-level AI Engineer position on the core product team at an early-stage AI startup, building agentic systems that automate complex, multi-step workflows across regulated and enterprise domains. You'll work across the full stack to ship production LLM-based services, with direct influence on product direction and user impact.
What You'll Do- Design, build, and maintain agentic systems that automate multi-step workflows across domains such as healthcare, legal, fintech, logistics, and compliance.
- Own production retrieval-augmented generation (RAG) pipelines and retrieval infrastructure - including vector DBs, embeddings, and indexing - for domain-specific search at scale.
- Implement multi-agent orchestration, tool-calling, memory, and reasoning components to deliver robust AI-driven user experiences.
- Develop evaluation and safety infrastructure to measure model performance, surface regressions, and enforce enterprise-level trust and reliability.
- Ship full-stack AI products from MVP to enterprise-grade, designing APIs and data models, writing frontend and backend code, and operating production systems with CI/CD, monitoring, and testing.
- Collaborate closely with founders, product, and design to prioritize work, define success metrics, and iterate on user feedback and telemetry.
What We're Looking For- 2-8 years of software engineering experience with a track record of shipping user-facing or backend products.
- Hands-on experience deploying LLMs or LLM-based services in production, including prompt design, orchestration, and tool integration.
- Proficiency across the stack: Python plus TypeScript/React (or equivalent), cloud platforms (AWS or GCP), and relational or NoSQL databases.
- Working knowledge of RAG patterns, vector databases, embeddings, and retrieval pipelines, with sound judgment on when and how to apply them.
- Experience building automated tests, evaluations, and monitoring for AI systems to ensure reliability beyond demo environments.
- Experience designing API-driven, high-throughput systems and real-time product features.
- Familiarity with agent or workflow frameworks (e.g., LangGraph, CrewAI) and orchestration tools (e.g., Temporal, Trigger) is a plus.
- Background in multi-tenant or enterprise-ready systems, or experience in regulated industries (healthcare, fintech, legal), is a plus.
- Familiarity with fine-tuning, parameter-efficient tuning, or multi-modal model integration is a plus.
- Strong ownership mindset - comfortable driving features end-to-end from data model to deployment and monitoring.
Compensation & BenefitsSalary range:
$180,000 - $400,000 USD annually. Visa sponsorship is not available.
LocationOn-site in
San Francisco, CA, United States.
