Senior Software Engineer, Internal AI

ABOUT ZIP Zip is the AI platform for enterprise procurement — built for humans and agents working together. By orchestrating procurement across teams, tools, and suppliers with the help of AI agents, companies can secure the resources they need to innovate faster than ever before. The world’s most influential enterprises trust Zip, including T-Mobile, OpenAI, AMD, Mars, Dollar Tree, and more. Together they’ve saved over $8 billion and processed over $500 billion in spend. Zip’s team includes product leaders from Apple, Airbnb, and Meta, as well as former procurement leaders from United Health, Sanofi, MGM Resorts, Discover, and NASA. Backed by Adams Street, Alkeon, BOND, CRV, DST, Tiger Global, and Y Combinator, Zip has raised $371 million, most recently at a $2.2 billion valuation and has been recognized by Forbes Fintech 50, Fast Company's Most Innovative Companies, Inc. Best in Business, and LinkedIn Top Startups. YOUR ROLE You are the engineering anchor of Zip’s new Internal AI team. The Internal AI team operates in a hub-and-spoke model: we own a central roadmap of internal product and platform capabilities that accelerate AI adoption across every non-software engineering use cases at Zip and we partner with those teams so their own roadmaps ship safely and well. This role is the technical anchor of that hub. You will build central capabilities used across the org, serve as the architecture funnel for non-product AI, and operate as a central interface into Zip’s core Engineering organization — so the rest of the business can move fast without creating chaos for Foundation and Security. This is an application engineer with infrastructure sensibility. YOU WILL - Build central platform capabilities that multiply the rest of the org. Ship the services, agents, centralized products, and MCP integrations that non-software engineer teams use as building blocks — the hub in our hub-and-spoke. - Be the architecture funnel for non-product AI. Review use cases coming out of every non-product team. Recommend the right path — off-the-shelf tool, Claude skill, Notion agent, no-code (n8n / Workato), or custom build. Squash the large majority of low-value vibe-coded ideas before they consume engineering or security cycles. - Hold the application-layer bar. You set and enforce application-level standards: integration patterns, data handling, identity, evals, observability. - Be the single interface into core Engineering. When something genuinely needs infrastructure — hosting, production scheduling, new data in the warehouse — you are the funnel point from the business into Foundations and Security. Triaged, reduced, sequenced, spec’d, so other teams aren’t each having the same conversation three times. - Ship when building is the right call. When no off-the-shelf path fits, you build — to production standards. Reviewed, version-controlled, observable, documented, owned. Never orphaned. QUALIFICATIONS - 6+ years building and operating backend applications in production. Bias toward application, integration, and data-pipeline work rather than pure infrastructure. - Real infrastructure sensibility. You understand hosting, identity, secrets, and data flow well enough to collaborate credibly with a core infrastructure team. - Strong architectural vocabulary. Comfortable reviewing others’ designs (engineers and non-engineers) and saying “use this pattern, not that one” with empathy. - Pragmatist. Picks the smallest tool that works — sometimes Python/Node, sometimes n8n, frequently “don’t build it, here’s an existing skill you should use with Claude Cowork” - Hands-on with modern AI tooling: Claude Code, Cowork, MCP, LLM APIs, evals. Opinionated about where AI is the wrong solution. - Strong communicator. Much of this role is consulting engineers and non-engineers and holding a line without becoming a blocker. - Comfortable in ambiguity. The function is new and the paved roads don’t fully exist yet. NICE TO HAVE - Experience sitting between a c