Ramp Policy Agents

News coverage:
Fast Company
CNBC

Ramp Policy Agents replaces manual and repetitive finance tasks with automations referencing natural language documents to make accurate company-specific judgement calls.

For example, instead of approving or denying company expenses one-by-one, Ramp uses natural language to capture what is and isn’t in policy to automate expense approvals with more than 99% accuracy. This frees up the finance team’s time for more strategic jobs, ensures spend is within the company rules, and gets expense approvals to employees instantly.

I was the Staff Product Designer leading this project, managing 2 designers, conducting weekly research, while collaborating with a team of engineers and PM.

Our team launched natural language expense policy in 2026, it was adopted by enterprise finance teams at Airbnb, SpaceX, Figma, and more. The feedback has been highly positive, some companies remarking that they have been able fully automate expense reporting and reducing headcount on teams to 0.