Chief financial officers have spent two years saying they do not trust AI with their numbers. Hyphenate raised $45M on the argument that they are wrong, and that its software can prove it.
Nexus Venture Partners led the Series A, joined by B Capital and returning backers Foundation Capital, BoldCap and Gaia Ventures. Total funding reaches $54M. The company, formerly Maximor, runs AI agents across general ledger accounting, treasury, order-to-cash, procure-to-pay and reporting.
The pitch avoids replacing the enterprise resource planning stack. Agents plug into existing accounting, banking and ERP systems and learn the work already happening inside them.
Tribal knowledge, not prompts
Instead of having finance staff author rule sets, Hyphenate trains agents on years of spreadsheets, invoice trails and email threads. That history, according to co-founder and chief executive Ramnandan Krishnamurthy, carries the institutional memory of a finance department.
Every autonomous action is logged for audit. When a human overrides an exception, the agent treats the decision as precedent.
The trust curve matters because the fear is real. In Hyphenate’s own survey, 96 percent of finance leaders saw value in automating repetitive work but only 14 percent trusted AI to do it unsupervised, and 97 percent wanted a human check.
Reference customers do the selling
Six public companies use the platform, including Groupon, Synthesia, HiBid and Mini Melts USA. Hyphenate says clients removed up to 90 percent of routine manual tasks and cut audit exceptions by about 75 percent.
HiBid pushed more than $2.5B in gross merchandise value through the system and shrank month-end close from three weeks to a single day.
Customer count grew elevenfold and revenue 86-fold over the past year, though the company did not give absolute figures. The next target is forecasting and capital allocation.
