Financial Operations Platform Ramp has debuted its first AI (AI) agent.
The new product is designed for controllers, and is the first of a set of agents scheduled to be released this year, which will automatically enforce the company's expense policy, block fraud, and help to stop fraud.
“The less the financial team is asked to do more, but the features remain primarily manual,” Lamp said in the release. “Today, teams using legacy platforms spend up to 70% of their time on tasks like cost reviews, policy enforcement, and compliance audits. As a result, 59% of controllership role experts report that they will generate some errors each month.”
Ramp says that controller-centric agents will solve these problems by eliminating redundant tasks, autonomously monitoring costs, working autonomously to enforce policies, applying “human-like in context-aware” inferences, and managing the entire workflow independently.
“Unlike traditional automation, which relies on basic rules and conditional logic, these agents act on behalf of their finance teams, implement large spending policies, prevent violations immediately, and strive independently to continually improve company spending guidelines,” the release added.
Pymnts wrote this week about the “Agent AI Promise” that not only generates or analyzes content, but also moves to start decisions, workflows beyond passive tasks and interacts with other software to complete projects.
“It's not just the brain, it's the agency, it's AI,” the report said.
Industry such as finance, logistics and healthcare use these tools to book meetings, process invoices, and autonomously manage the entire workflow.
However, while some corporate leaders may retain the lofty view of autonomous AI, the latest PYMNTS intelligence in the June 2025 CAIO report shows that “AI at a Crossroads: Agent Ambition is Full of Operational Reality” shows gaps in trust between executives when it comes to agent AI, which highlights serious concerns about accountability and compliance.
“However, full-scale companies still remain limited in recruitment,” writes Pymnts. “Despite increasing capabilities, Agent AI is deployed in experimental or limited pilot settings, with the majority of the system operating under human supervision.”
But why are middle market companies worried about harnessing the power of autonomous AI? The answer is strategic and psychological, Pymnts adds, noting that while the technical potential is huge, the preparation of the system (and human) is very troublesome.
“In order for AI to act autonomously, executives must trust not only the output, but the entire decision-making process behind it. That trust is difficult to gain and easy to lose, and the study stated that “80% of high-automatic companies cite data security and privacy as the highest concerns of agent AI.”
