Ramp's policy agent reviews expenses against company policy for 1,000+ finance teams
Ramp launched AI agents for controllers that read a company's expense policy, approve low-risk spend, flag suspicious receipts, and answer employee policy questions, with an audit trail for each decision.
- Organization
- Ramp
- Industry
- Fintech
- When
- Jul 2025
- Evidence
- Vendor case study
- Agents and tools
- Ramp Policy AgentOpenAI reasoning models
Reported outcome
Used by 1,000+ finance teams within six months; Ramp claims 99% approval accuracy and 7x more out-of-policy spend caught.
- 1,000+Finance teams
- 99% (vendor)Approval accuracy
- 7x (vendor)Out-of-policy caught
Caveats
Published by the company that sells the agent. Treat numbers as marketing until confirmed elsewhere.
- All performance figures come from Ramp.
- 'Accuracy' is measured against Ramp's own labels, method not published in the source.
What the agent does
The agent reads the company's written expense policy. For each expense, it approves low-risk items or sends a recommendation with reasons to a human approver. It flags suspicious receipts and invoices and answers employee questions about policy.
Why this works
Expense review is high-volume and rule-based, with some judgment. An agent that explains each decision gives finance teams something they can audit.
What to copy
- Write your policy as clear rules first. The agent can only enforce what is written.
- Auto-approve only low-risk categories at first. Widen as trust grows.
- Keep the reason for every decision, for audits.
Starter prompt
Written by us, not by Ramp.
Expense policy:
{policy_text}
Expense: {merchant}, {amount}, {category}, {date}, memo: "{memo}",
receipt text: "{receipt_text}"
Decide: APPROVE, REVIEW, or REJECT.
Quote the exact policy line that supports your decision.
Choose REVIEW if the receipt is missing, the amount differs from the
receipt, or the policy does not clearly cover the case.
Sources
- Ramp Policy Agent is generally available — Ramp, Jan 2026
- Ramp introduces AI agents suited for controllers — CPA Practice Advisor, Jul 2025
Added Oct 16, 2026. Spotted an error? Send a correction.
Get new use cases weekly
One short email. Sources included.
Related use cases
BNY runs 100+ AI 'digital employees' with their own logins and human managers
Over 100 digital employees in production by October 2025, later reported as 130+, each with a human supervisor.
Morgan Stanley's Debrief agent takes notes in advisor meetings and drafts the follow-up email
Rolled out to the firm's financial advisors in 2024; the firm cites about 30 minutes saved per client meeting (testimonial).