What you'll do Build the surfaces agents call directly: API-first account provisioning, machine-to-machine authentication, scoped credentials, and agent-level identity and attribution. Own our MCP (Model Context Protocol) servers and an agent-first CLI — let coding agents and LLMs discover and invoke Cloudinary reliably, engineered for agentic callers: token-efficient tool schemas, machine-parseable output, and loud, actionable failures instead of silent ones. Harden the agent media path — make upload and core media operations reliable for non-human callers, eliminating the ambiguity and silent-failure modes that cause agents to drop off mid-task. Make agent behavior measurable — partner on the instrumentation, signals, and attribution that tell us whether agents are succeeding. Work across a polyglot backend — Node.js/TypeScript for the agent tooling layer, with Ruby/Rails and Go in the core platform. Ship, measure, iterate — prototype quickly, dogfood the experience, and represent the agent's perspective in every design decision. Work closely with Product, Ecosystem, and Discoverability across the AX team. About you 10+ years building and operating production APIs, SDKs and Backend systems and at scale. Strong in Node.js/TypeScript, and comfortable working in (or quickly ramping on) Ruby/Rails and/or Go. Deep API and protocol design fundamentals — authentication (OAuth, API keys, scoped permissions), versioning, idempotency, rate limiting, clean contracts. Genuine curiosity about agents and LLMs and how they're reshaping how software gets built and run. Solid distributed-systems and cloud instincts (AWS or equivalent). A bias for ownership in ambiguity — you scope the problem, not just the solution, and you ship. Strong written communication — it matters more than usual when your consumers are machines and your teammates are distributed. Advantages Hands-on experience with MCP, LLM tool/function-calling, or building AI-/agent-native products. Experience building SDKs, CLIs, or developer tools that other engineers depend on. Background in media, imaging, or video processing/delivery. Open-source contributions in the agent or developer-tooling ecosystem.
