Description

You will architect the enterprise Gemini platform, including managed chat, enterprise search, agent galleries, and custom front-end applications.

This role is hybrid.

Responsibilities

• Design and govern the Agent Platform layer, including ADK, agent runtime, registry, identity, security, and observability.
• Architect integration and data layers across Google Workspace, Microsoft 365, MCP servers, and federated retrieval across function-scoped data stores.
• Manage identity and access patterns, including OAuth2, SAML, RBAC, token exchange, and document-level ACL mapping.
• Drive cost awareness across token and inference spend as the agent estate grows.
• Establish AI governance functions, including review boards, architecture checklists, and audit evidence.

Required Skills

• 10+ years of experience with a Bachelor's degree.
• Google Gemini Enterprise and ADK (or comparable enterprise agent platform).
• Multi-agent systems in production: orchestration, routing, tool use, memory, human-in-the-loop.
• RAG architecture: vector stores, embedding models, chunking, hybrid search.
• Identity and access: OAuth2, SAML, RBAC, token exchange, service-account and end-user credential propagation.
• Python (proficient); Go or equivalent second language (comfortable).
• MCP: building servers, not only consuming them.
• GCP: Cloud Run, GKE, Vertex AI, networking, IAM.
• Containerisation and orchestration: Docker, Kubernetes.

Preferred Skills

• Production deployment of agents on Gemini Enterprise / Agent Platform.
• AI evaluation tooling: Langfuse, LangSmith, Braintrust, or custom eval frameworks.
• Multi-model routing across Gemini, Claude, OpenAI, or Llama.

Key Skills

Education

Bachelor's degree