Ramp targets KPMG rivals with AI operating system for accounting firms

Ramp’s new Stack product is aimed directly at the work that junior accounting staff and firm operations teams have long absorbed: bookkeeping, reconciliations, month-end close, journal entries, variance analysis, flux analyses and client workflows. By rolling those tasks into what Ramp calls an AI operating system for accounting firms, the company is signaling that the next wave of competition is not about chatbots or point tools, but about taking over the daily mechanics of firm production.

Ramp launched Stack on June 3 as its first product built specifically for accounting firms, and it is already leaning on scale to make the pitch. The company says it works with more than 4,500 accounting firms and that 92 of the top 100 CPA firms already have clients on its platform. Ramp also says the accounting-firm market is roughly $150 billion, and that Stack will be free through August, a move that looks designed to win a place in firm workflows before pricing becomes a factor.

The labor message is hard to miss. Ramp says more than 300,000 CPAs have exited the profession and accounting degrees hit a 20-year low last year, which helps explain why firms are turning away clients they cannot staff. The company’s design-partner program started with a first cohort of more than 20 firms, including fractional controllers, outsourced bookkeeping shops, virtual CFO firms and regional CPA practices. That mix suggests the product is not being sold just to large firms, but to the middle layers of the profession that often carry the heaviest volume work and the thinnest staffing.

For KPMG and its rivals, the bigger question is what gets compressed and what gets added back as oversight. Ramp says Stack includes human sign-off, real-time visibility into every reason, source and step, and an auditable workflow. That kind of promise can reduce time spent assembling workpapers, chasing down reconciliations and cleaning up messy books, but it also shifts effort toward review, exception handling and control validation. In other words, the firm may need fewer people pushing routine transactions, but more people policing whether the machine is behaving the way clients, regulators and engagement partners expect.

That shift lands squarely in a market where KPMG is already telling clients that finance is moving fast. In May, KPMG said 93% of U.S. companies will be deploying or scaling AI in their finance functions in the next 18 months, with half planning multi-agent systems across workflows. KPMG’s 2026 Global AI in Finance report also found that only 23% of organizations say AI is exceeding expectations, a reminder that the hard part is no longer trying AI, but making it reliable enough to sit inside real operating processes. Ramp is trying to sell exactly that: not another assistant, but a system that captures procedure, makes decisions reviewable and bakes controls into the workflow from the start.