Applied AI Engineer

About Warp

We process $X billion in annual payments for high-growth companies across all 50 US states + globally.

We've raised $85M from Battery Ventures, Sapphire Ventures, Peak XV, Sound Ventures (Ashton Kutcher), and Y Combinator, alongside operators including Tobi Lütke (Shopify), Claire Hughes Johnson (Stripe), and the founders of Dropbox and Replit. Our team comes from MIT, Ramp, Brex, Google, Gusto, Rippling, and more.

About the Role
High-growth companies spend thousands of hours on people operations: onboarding new hires, running payroll, managing benefits, provisioning devices, answering policy questions, approving time off and expenses. Most of that work still runs on spreadsheets, tickets, and people chasing people. We're building agents that do it for them.
Warp Agent works in the background for HR, finance, and people teams. It gets new hires ready for day one, reviews expenses against policy, handles time-off requests, checks payroll before it runs, and reports on what's happening across the company. Customers hand it a receipt or a notice and it takes care of the rest. Because Warp runs the full employee lifecycle on one platform, the agent has the context to finish the work, not just suggest it.
The hard part is trust. People data is the most sensitive data a company has, and people operations leave little room for error. Every agent action respects role-based permissions, so an IT manager's agent can provision a laptop without ever seeing anyone's salary, and every action is logged and auditable.
We believe agents will soon do more work on Warp than people do. As an Applied AI Engineer, you'll build those agents, the runtime they operate on, and the systems that make them reliable enough to run a company's people operations.

What You'll Do

• Ship agentic features end to end, from the model call to the UI, across HRIS, payroll, benefits, IT, and more
• Build and extend Warp Agent's harness: tool-calling, durable execution for long-running work (triggers, scheduling, wait and resume), and a run log that makes every agent action auditable
• Design evals and guardrails for workflows where mistakes are costly, and decide when an agent acts on its own versus asks for approval
• Build new agent workflows and expand what customers can configure and author themselves
• Extend our permission model to agents so every action respects who the agent is working for
• Build internal agents and tooling that remove manual work across Warp's own teams
• Build new surfaces for interactions with Warp agent across the web app, employee app, mobile app, and slack
• Collaborate directly with the CTO and CEO

You'll Be Good for This Role If You

• Have shipped LLM features that real users depend on in production, and know where they break
• Think in systems: evals, observability, cost, latency, and failure modes
• Are product-minded and care how an agent's work shows up for the person who delegated it
• Get excited by hard domains where correctness matters more than demos
• Use coding agents heavily in your own workflow and have opinions on what works

What You Need

• Strong full-stack engineering skills and expertise in a modern, strongly typed language
• A track record of building production AI systems: agents, tool use, retrieval, or structured extraction
• Experience designing reliable backend systems
• High agency. You figure things out.
• Willingness to work from our NYC office five days a week

Benefits at Warp:

Warp is based in New York City, and most of our team works together in-office 5x per week. Please refer to the top of this page for this position's hiring location and working model.