AI Platform Engineer
Core
Build the execution layer for Supabase's internal AI systems, including an event-triggered queue, a headless model-agnostic runtime, durable state, human review gates, and a comprehensive evaluation layer to ensure agent trustworthiness.
Role type
Staff AI Platform Engineer
Builds
Internal AI operating system, agent platform, evaluation layer, and agent portfolio (reporting, drafting, triage, Q&A agents)
Domain
AI/LLM operations, internal enterprise systems, cloud infrastructure
Deliverable
production ML models | infrastructure
Required skills
shipping production LLM agent systems, designing evaluation frameworks with golden suites and behavioral assertions, deep API work and MCP server authorship, end-to-end infrastructure ownership in Python on GCP, cloud warehouse management, infrastructure as code
Preferred skills
public work in agent frameworks or MCP servers, LLM observability and cost instrumentation
Technologies
Python, GCP, cloud warehouse, infrastructure as code, MCP servers
Responsibilities
Ship the agent platform to production with event-triggered queues and model-agnostic runtimes; own the evaluation layer with CI gates; build and register the agent portfolio; enforce governance in code via risk tiers and tool-permission gates; design system contact protocols; own platform tooling and compute operating measures from production data
Seniority
Staff, hands-on IC with architectural ownership
