- Jobs
- Varick Agents
- AI Engineer
AI Engineer
Full-timeMid
$220K - $250K/yr
AI Tools
Claude CodeCrewAILangGraphMCPRAGVector Databases
Tech Stack
PythonLangGraphCrewAIQdrantpgvectorPineconeMCP
Agent Workflow
Design agent architectures for complex workflows with multi-step reasoning and tool calling. Build evaluation systems for agent quality. Engineer prompt systems, retrieval pipelines, and context strategies. Develop feedback loops enabling agents to learn from human corrections.
About the Role
About Varick Agents
Varick deploys large-scale AI transformations at major enterprises, automating entire departments through interconnected agent systems integrated into existing infrastructure.
AI Engineer - San Francisco (SF-based required) Full-time | Posted March 22, 2026
Compensation: $220,000-$250,000/year with equity
Benefits: 100% medical, dental, vision; MacBook Pro + peripherals
Key Responsibilities:
- Design and implement agent systems for complex enterprise workflows involving multi-step reasoning and exception handling
- Develop evaluation systems measuring agent quality, accuracy, safety, and factual grounding
- Engineer prompt systems and retrieval pipelines ensuring reliable agent behavior
- Create feedback mechanisms enabling agents to learn from human corrections
- Optimize inference costs and latency for production deployments
- Establish governance and observability standards for agent reliability
Requirements:
- 3+ years software engineering experience with 1-2 years focused on LLM or AI systems in production
- Hands-on experience building agent workflows with tool integration and multi-step reasoning
- Deep expertise in prompt engineering and context optimization for LLM reliability
- Experience building evaluation frameworks for AI outputs
- Strong Python and backend engineering fundamentals
- Demonstrated shipping of AI features to real users, managing hallucinations and edge cases
Helpful Experience:
- Frameworks: LangGraph, CrewAI, Claude patterns
- Vector databases: Qdrant, pgvector, Pinecone
- MCP, tool-calling protocols, API integrations
- Fine-tuning, LoRA, model adaptation
- Enterprise AI deployments with governance and compliance