Dive Brief:
- San Mateo, California-based Maxima is rolling out a new AI agent dubbed “Max” to add to its accounting platform that works with a company’s existing enterprise systems, according to a Wednesday press release.
- The AI accounting agent can handle recurring work typically done manually that includes payroll entries, balance sheet reconciliations and variance explanations, which is then routed back to accountants for approval before being posted to the general ledger, the company states.
- The new product combines in one agent capabilities for transactional accounting, month-end workflows and close processing, as well as accounting intelligence, three areas that were previously handled by multiple agents, according to Yogi Goel, co-founder, CEO and CFO of Maxima. “Rather than interacting with individual agents or modules, accountants can now delegate prep work across multiple use cases directly to Max, just like a teammate,” Goel said in an emailed response to questions.
Dive Insight:
Maxima’s platform update comes as CFOs and accounting teams are navigating AI and other tech-related changes now reshaping how companies handle basic accounting and finance tasks. AI-enabled accounting is expected to reduce accounting costs and labor while also requiring the next generation of accountants to learn to make judgment calls on processes without the benefit of previous hands-on experience, CFO Dive previously reported.
Founded in 2024, the privately-held Maxima is backed by the venture capital firms Redpoint Ventures, Kleiner Perkins and Audacious Ventures, according to the release. The company bills itself as an agentic system of work for accounting that is layered over existing enterprise resource planning systems.
Before the Max update, accountants “still had to drive the process themselves,” Goel said. For example, they still had to gather data, build schedules and prepare journal entries. With Max the company said early users of it are saving as much as 60 hours of labor per employee every month. It differs from AI chatbots, he says, in that it can “orchestrate muti-step workflows.”
“Unlike general purpose AI systems such as Claude or ChatGPT, Max is not a standalone agent layered on top of the general ledger,” Goel said. “It was designed specifically for accounting, with audit trails, role-based permissions, segregation of duties, validation checks, deterministic execution, human-in-the-loop review and approvals, and domain-specific accounting tools.”
Josh Waldron, CAO and interim CFO of Scale AI in San Francisco which provides data infrastructure support to large language model labs, said his company began using Maxima in late 2024 to handle reconciliation of outward payments, a process that is now instantaneous. Previously it could take as long as two weeks as the company had to manually sort through data from numerous cash accounts around the world and assess such variables as currency fluctuations.
Waldron says his firm benefits from Maxima’s ability to connect to various systems, pull the data, and automatically upload into Scale AI’s Netsuite ERP system. In contrast, Waldon said some tools require companies to download bank statements and then put it into the system, which is a laborious process. Since that change, Waldron said he gained confidence in the system and went on to overhaul other elements of the companies’ close process, Waldron said.
“It would take us at times the entire month in order to close the books and this last month we closed in the first two weeks of the month which is a record for us,” Waldron said.
To date, Maxima has raised $41 million in funding and processed transaction volume valued at over $400 billion.
