Anthropic Claude AI Manager -- Direct client

Job Title: Anthropic Claude AI/Data Manager
Location: Remote/SFO, CA
Duration: Contract to Hire
Client: Direct

Required Qualifications
Experience
• 5-8 years combining technical AI/data engineering AND client-facing consulting or advisory roles. Both are non-negotiable.
• Demonstrated track record delivering AI solutions in production -- can describe a live deployment, its architecture, failure modes, and lessons learned.
• Prior experience in financial services, wealth management, fintech, or an adjacent regulated industry.
• Experience working directly with senior business stakeholders: can write the exec summary AND debug the API call in the same day.

Anthropic / Claude -- Must-Have Skills
• Hands-on, production experience with the Anthropic Claude API. You have built something real with it.
• Proficient in Claude system prompt engineering: context window management, persona design, tool use (function calling), multi-turn architecture, and prompt injection defense.
• Experience configuring Claude Enterprise: org API key setup, ZDR enablement, DPA implications, Sonnet vs. Opus trade-offs, and rate limit management.
• Understanding of Claude''s Constitutional AI approach and how to communicate it to compliance-minded financial services clients.
• Familiarity with Anthropic''s MCP (Model Context Protocol) and at least one integration using MCP connectors.

Technical Engineering Skills
• Python -- proficient and production-ready. Code is readable, documented, and tested. Not just Jupyter notebooks.
• Git -- all code version-controlled, commits are meaningful, work is reviewable via PR.
• RAG architecture: experience building retrieval-augmented generation pipelines with vector databases (Pinecone, Chroma, Weaviate, pgvector, or equivalent).
• Agentic AI: multi-step, tool-using agents with LangChain, LlamaIndex, AutoGen, or equivalent -- including memory, state management, and error handling.
• REST APIs and data integrations: connect LLM pipelines to real data sources with proper OAuth, API key management, and security hygiene.
• Cloud platform exposure: AWS, Azure, or Google Cloud Platform for deploying and managing AI workloads.