Stanford's Enterprise Technology team needs to design, implement, and support AI solutions across university use cases, specifically focusing on LLM/RAG services and integration with enterprise platforms.
Requirements
- Agent/Agentic Framework Experience: Built and shipped at least one production LLM agent or agentic workflow using frameworks such as LangGraph, LangChain, CrewAI/AutoGen, Google Agent Builder/Vertex AI Agents (or equivalent). Able to explain tool selection, orchestration logic, and post‑deployment support.
- Proven Delivery: Implemented 1+ AI/ML projects and 1+ GenAI/LLM projects in production, with operational support (monitoring, tuning, incident response). Projects should serve sizable user populations and demonstrate measurable efficiency gains.
- Strong understanding of AI/ML concepts (LLMs/transformers and classical ML) and experience designing, developing, testing, and deploying AI-driven applications.
- Programming Expertise: Proficient in Python; familiarity with Node.js/TypeScript/React and RESTful APIs; ability to read/extend existing codebases.
- Vector & Search Basics: Worked with at least one vector/search tech (e.g., Pinecone, OpenSearch/Elasticsearch, FAISS, Milvus) and basic embedding workflows.
- Experience with cloud AI stacks (e.g., Google Vertex AI, AWS Bedrock, Azure OpenAI) and vector/search technologies (Pinecone, Elastic/OpenSearch, FAISS, Milvus, etc.).
- Thorough understanding of SDLC, MLOps, and quality control practices.
Responsibilities
- Assess user needs and requirements,
- Turn requirements and tickets into well-engineered components (data prep, pipelines, vector stores, prompts/agents, evaluation hooks).
- Build, maintain, and update programs like LLM-based agents/services that securely call enterprise tools (ServiceNow, Salesforce, Oracle, etc.) using approved APIs and tool-calling frameworks.
- Create lightweight internal SDKs/utilities where needed.
- Configure and optimize RAG workflows (chunking, embeddings, metadata filters) and integrate with existing search/vector infrastructure—escalating architecture changes to designated architects.
- Contribute tests, CI/CD pipelines, telemetry, and prompt/model versioning; participate in code reviews and release activities across dev/test/prod; follow team software development methodology.
- Apply established guardrails (PII redaction, policy checks, access controls/minimum-privilege).
Other
- Bachelor's degree and three years of relevant experience or a combination of education and relevant experience.
- Ability to define/solve logical & technical problems for highly technical applications; strong problem-solving and systematic troubleshooting skills.
- Excellent communication, listening, negotiation, and conflict resolution skills; ability to bridge functional and technical resources.
- May work extended hours, evenings, and weekends.
- Interpersonal Skills: Demonstrates the ability to work well with Stanford colleagues and clients and with external organizations.
