Sr. Distinguished AI Engineer (agentic AI Platform)

Capital One Capital One · Banking · San Jose, CA +4

Senior Distinguished AI Engineer focused on building and scaling an Agentic AI Platform at Capital One. The role involves contributing to platform architecture, standardizing agentic workflows using frameworks like LangGraph and AutoGen, developing GenAI SDKs/CLIs, implementing central guardrail services for trust and safety, optimizing orchestration for cost reduction, and driving innovation in areas like multimodal RAG and hierarchical agent memory. The role also includes coaching and evangelizing the platform vision.

What you'd actually do

  1. You will contribute to the north star platform architecture, continuously publishing and refining living diagrams and canonical APIs that cover agent orchestration, RAG pipelines, prompt libraries and multi-tenant policy enforcement.
  2. A major emphasis is around standardizing and automating agentic workflows : you will evaluate agentic frameworks in LangGraph, AutoGen, Semantic Kernal, CrewAI and LlamaIndex and then harden / blend patterns that best meet enterprise SLAs do that 90% of new apps adopt them.
  3. Developer experience is another cornerstone. You will contribute to crafting an end to end GenAI SDK, CLI and starter kits that let AI engineers spin up secure, observable agentic workflows in under minutes, shrinking prototyping to production timelines by 30%.
  4. Trust and safety remain paramount; you will help bring together a vision of central guardrail services - prompt firewalls, content-filter hooks, red team harnesses and audit APIs - consumed by every application to ensure zero Sev4 incidents.
  5. You will collaborate with cross organization architects to drive end to end performance by optimizing orchestration - level batching, retrieval caching, heuristic tuning to achieve reductions in per token spend.

Skills

Required

  • Python
  • Go
  • Scala
  • Java
  • Cloud platforms (AWS, Google Cloud, Azure)
  • Agentic Frameworks (LangChain, CrewAI, Semantic Kernel, AutoGen)
  • LLMOps (Vertex AI, SageMaker, Azure ML)
  • AI systems design and development
  • Machine learning platforms

Nice to have

  • C++
  • C#
  • Golang
  • Helm charts
  • Operators
  • CRDs
  • Hierarchical agent memory
  • Multimodal guardrails
  • Multimodal RAG
  • Similarity Search and VectorDBs
  • Guardrails
  • Memory

What the JD emphasized

  • enterprise SLAs
  • 90% of new apps adopt them
  • shrinking prototyping to production timelines by 30%
  • zero Sev4 incidents
  • reductions in per token spend

Other signals

  • building agentic platforms
  • standardizing agentic workflows
  • enterprise SLAs
  • developer experience for GenAI SDK/CLI
  • central guardrail services
  • optimizing orchestration
  • multimodal RAG
  • auto scaling agents