New startup Zalos raises $3.6 million to automate complex finance tasks using AI agents that mimic human screen activity.

Key points

• Solving the "Integration Gap": Many finance teams struggle because their various software tools (like SAP, NetSuite, and Excel) don't "talk" to each other, forcing employees to manually move data between them.
• No APIs Required: Instead of building complex technical connections between software, Zalos uses AI agents that learn by watching screen recordings of employees performing their actual work.
• Human-like Operation: The agents can log into existing systems, navigate menus, handle two-factor authentication, and perform data entry just like a human analyst would.
• Finance-Specific Focus: Unlike general-purpose AI, Zalos is built specifically for the high-accuracy and audit-heavy requirements of corporate finance departments.
• Strong Backing: The company is part of the Y Combinator Fall 2025 cohort and has attracted investment from high-profile finance executives, including the CFOs of FedEx and Tide.

By bypassing the need for expensive and time-consuming software integrations, Zalos could significantly speed up how large companies handle back-office tasks like month-end closing and billing. It represents a shift toward "computer-use" AI that works within existing systems rather than requiring companies to replace their entire software stack.