Finance departments have spent years layering enterprise software on top of one another, only to find that many core processes still depend on staff moving data manually across disconnected systems. Zalos, a startup building AI “computer agents” for finance operations, says it wants to automate that work without requiring companies to replace the systems they already use.
The San Francisco-based company said on March 24 that it had raised $3.6mn in seed funding. The round was led by 14 Peaks, with participation from Cohen Circle, 20VC and a group of angel investors.
Zalos is developing software agents designed to log into finance systems, navigate screens, enter data and carry out defined workflows across tools such as ERP software, Excel, email and internal platforms. The company says the approach is intended to address a longstanding problem in finance operations: many systems used for billing, reconciliation and reporting do not connect cleanly, leaving finance teams to bridge the gaps themselves.
Rather than relying on extensive integrations, Zalos says its software is trained using screen recordings of finance workflows. Once trained, the agents are intended to reproduce those steps inside existing systems. The company says the platform currently works with NetSuite, Sage and SAP S/4HANA, and is being used in areas including billing automation, month-end reconciliations and KPI reporting across multiple ERP instances.
The company argues that finance presents a distinct challenge for AI automation because the tolerance for error is low and the need for traceability is high. Zalos says its platform records agent actions in auditable logs and includes enterprise controls such as SOC 2 Type II certification, single sign-on, role-based access controls and on-premise deployment options.
The funding comes as AI companies push beyond chatbots and copilots into software designed to perform actions directly inside digital systems. Zalos says finance is a particularly attractive target because ERP environments are deeply embedded, expensive to replace and often still generate large volumes of repetitive manual work.
William Fairbairn, Zalos’s chief executive and co-founder, said the company was built on the view that finance leaders should be able to adopt AI without overhauling the software stack around which their operations already run. Hung Hoang, the group’s co-founder and chief technology officer, said the company sees agent-based automation as an alternative to both large-scale ERP replacement and API-dependent integration efforts.
Zalos was founded by Fairbairn and Hoang after the pair concluded, from different professional backgrounds, that finance teams remained constrained by legacy system architecture. Fairbairn previously worked at Agicap, where he said finance executives often described ERP implementations as slow, costly and risky. Hoang, who previously worked at Apple Pay, became focused on automation approaches that could function without relying on APIs.
The company said it plans to expand beyond the midmarket ERP systems it currently supports and move further into enterprise ERP and on-premise environments. It also aims to build a broader layer of contextual data across finance systems so that customers can deploy multiple agents across a wider set of workflows.
