Own the evolution of our shared Agent SDK and adjacent developer-facing libraries.
Define the default engineering patterns for agent runtime behavior, tool use, structured outputs, context management, retries, testing, and deployment.
Build the observability layer for AI systems, including tracing, prompt and model version visibility, tool-call telemetry, cost tracking, latency, failure modes, and fallback behavior.
Create practical dashboards, alerts, and operational workflows that let us catch regressions before the business feels them.
Lead the design of our experimentation and evaluation platform for model-backed and agent-backed systems.
Improve how we compare prompts, providers, retrieval strategies, model versions, and workflow designs.
Build greenfield internal tooling that accelerates company-wide AI adoption, especially around MCP-style tools, app-builder patterns, and reusable internal AI primitives.
Help define what great AI-enabled work looks like inside the company by modeling strong usage patterns and turning them into scalable defaults for other teams.
Turn fragile or manual workflows into well-instrumented platform capabilities with good defaults and clear ownership.
Partner with engineers and product teams to identify the highest-leverage shared investments, then drive them from design through production.
Raise the technical bar through architecture work, design reviews, implementation, mentoring, and strong engineering judgment.

You are a strong software engineer first, with staff-level judgment and a track record of owning technical systems that multiple teams rely on.
You have deep experience building platform, infrastructure, or developer tooling, ideally in environments where reliability and usability both matter.
You have experience with AI or ML platform problems such as agent runtimes, LLM tooling, inference infrastructure, experimentation systems, evaluation frameworks, or model observability.
You know how to design APIs, SDKs, and reusable abstractions that improve developer velocity without hiding important complexity.
You are an AI power user yourself. You actively use AI to accelerate engineering work, investigation, debugging, design, and knowledge work, and you have strong judgment about where it helps and where it does not.
You are comfortable operating in ambiguous, greenfield areas and can make pragmatic scope decisions without overbuilding.
You can work cross-functionally and lead through influence, clarity, and execution rather than title alone.
You enjoy being close to the code and architecture, even when operating at broad technical scope.

Staff Engineer, AI Platform

Key skills