Ramp expands its AI push into finance operations

Ramp has introduced a new service called Applied AI Solutions, a program that places its engineers inside customer finance teams to design and deploy custom AI workflows. The launch reflects a growing belief at the company that finance departments need more than generic software or model access to make artificial intelligence useful in day-to-day operations.

Ramp said the offering is aimed at organizations that want AI systems tied directly to finance processes such as capital planning, variance analysis, board reporting and financial close. According to the company, those tasks often depend on scattered information across systems, policies buried in documents or email threads, and institutional knowledge held by a few employees.

The announcement comes as AI spending rises sharply among Ramp customers. The company said token usage across its base of 70,000 customers has increased 13-fold since January 2025. Ramp also cited survey data indicating that while 87% of chief financial officers view AI as important, only 21% say it has produced measurable results.

A consulting-style service built around finance workflows

Applied AI Solutions is designed as a hands-on service rather than a self-serve product. Ramp engineers work directly with finance teams to map out operations, identify pain points and build tailored systems on top of the company’s platform. Ramp said the process includes connecting data from ERP systems, data warehouses, cloud storage and even paper-based workflows.

The company describes the core of the service as a “Finance Intelligence Layer,” a semantic framework that translates how a business actually works into structures that AI agents can use. That layer is intended to turn fragmented data into organized objects and attributes that reflect operational reality.

Ramp said the resulting AI agents can be built to interact with a company’s existing systems, reading and writing data within the tools finance teams already use. The company also said it is model-agnostic, meaning it benchmarks different AI models on real finance tasks and routes production workflows to the option it believes offers the best mix of performance and cost.

Built from Ramp’s own finance operations

Ramp said the new service grew out of tools it developed for its internal finance team. The company said it already relies on AI agents in production for work including financial planning, reconciliation and reporting. It said that internal setup allows its finance organization to operate with much less headcount than would normally be expected at its scale.

Ori Daniel, Ramp’s head of Applied AI Solutions, said in the company’s announcement that the service was shaped by lessons from its own use of AI as well as conversations with customers. He said the hardest part of applying AI in finance is not the model itself, but the work required to make a company’s data and business context understandable to agents.

Ramp said it plans to move quickly with each deployment, focusing deeply on a single workflow and shipping production systems within weeks. In some cases, the company expects to hand ownership of the solution to the customer after the initial build. In others, it said it can stay involved as the system expands.

The launch adds a services component to Ramp’s broader financial software offering and shows how the company is positioning itself as more than a spend management platform. With Applied AI Solutions, Ramp is betting that finance teams want embedded technical help to turn AI from a strategic priority into a working part of their operations.