Integrations Engineer

ABOUT RUNLAYER AI is transforming how every company operates, but most enterprises are stuck. They want to move fast with AI Agents, tools, and workflows, but they can't do it safely. We're fixing that. Our team built AI Actions for OpenAI, shipped Zapier Agents to millions of users, and launched the first remote MCP server with Anthropic. We helped establish the protocol, and now we're building the platform enterprises need to actually put AI to work. Runlayer is one platform for MCPs, Skills, and Agents: purpose-built security, fine-grained governance, and complete observability so organizations can go all-in on AI across the entire company without the risk. We just raised a $30M Series A led by Felicis, with participation from Khosla Ventures, bringing our total raised to $42M. Already trusted by Gusto, Instacart, Opendoor, dbt Labs, and Decagon. ABOUT THE ROLE As an Integrations Engineer, you'll own Runlayer's coverage surface: making every AI client, agent framework, and MCP server work first-class through the control plane. Coverage is the product, so you'll own integrations end to end, from design and launch through monitoring and debugging, and catch breaks before customers do. This is a hands-on, high-ownership role, and you'll help shape how the integrations function works as it grows. WHY YOU'LL THRIVE HERE - Impact: Coverage is the product; you'll make every AI client, agent framework, and MCP server work first-class through Runlayer. - Excellence: Work alongside engineers who helped establish MCP and ship fast. - Ownership: Own integrations end to end, from design and launch to monitoring, on-call, and incident response. WHAT YOU'LL DO - Own the full lifecycle of integrations: design, build, launch, monitor, and debug connectors across clients, agent frameworks, and MCP servers. - Make every major AI client (Claude Code, Claude Desktop, Cursor, Codex, Copilot, and the long tail) work first-class through Runlayer, from auth to policy enforcement, hooks, and session extraction. - Give agents built on any ADK (Claude Agent SDK, OpenAI Agents SDK, LangGraph, Google ADK, and whatever ships next) identity, governance, and observability, including agent accounts and attributing MCP calls to the agents that made them. - Keep 50+ third-party MCP connectors reliable: catch upstream breaks and auth regressions before customers hit them, and guide admins safely through migrations (URL changes, stdio-to-deploy, forced re-auth, scope changes). - Own reliability end to end: coverage smoke-tests on every release, monitoring, and fast debugging under real-time customer pressure. - Build the OAuth broker (the TypeScript/Express service managing OAuth 2.1 flows for third-party servers) and the tooling that scaffolds, validates, and publishes new integrations with little manual work. - Work cross-functionally with product, partnerships, and customers, and help shape how the integrations function grows. WHAT WE'RE LOOKING FOR - 5+ years building integrations, SDKs, or developer-platform connectors, with strong opinions on partner API design. - Expert-level OAuth 2.1: you've implemented authorization servers, token brokers, or complex multi-provider OAuth flows. - Agent-ecosystem fluency: you've built or operated MCP servers and shipped on at least one agent framework, and can explain transports, tool schemas, and session lifecycle from experience. - Strong TypeScript and Python. - You've built integration test suites that run against real third-party APIs with auth. - An operational mindset: you own systems in production, treat reliability as first-class, and can trace an issue from a customer report through logs and API responses to root cause. - You've maintained a catalog or marketplace of third-party integrations, with versioning, deprecation, and migration patterns. - AI-native workflow: you delegate real work to agents daily BONUS QUALIFICATIONS - Contributions to the MCP spec or to an agent framework. - Experience running or s…