Zalos Unleashes Computer Agents on the Finance Ops Grunt Work Nobody Wants to Do
Zalos, a Fall 2025 Y Combinator company, is taking direct aim at the repetitive, time-consuming, and error-prone work that still clogs finance operations at companies of every size. The startup builds AI finance automation software – computer-using agents that log directly into a company’s existing ERP, accounting systems, banking portals, and other CFO tools – then automate reconciliation, pricing, billing, and month-end close.
Founded in 2025 by William Fairbairn and Hung Hoang, Zalos is built around a simple observation: even with decades of SaaS innovation, finance operations is still drowning in grunt work. Legacy ERPs remain stubbornly resistant to modern AI. Many critical workflows live behind authentication-protected websites and portals that no integration vendor has ever touched. And most enterprises run dozens of finance systems that lack full integrations with one another. The result is a daily reality of swivel-chair work, late months, and finance teams too buried in transaction processing to provide real strategic insight.
What sets Zalos’s AI finance automation software apart is that it skips the integration problem entirely – agents automate processes based on simple human-friendly instructions and handle the messy realities of enterprise software natively, including 2FA prompts and CAPTCHA challenges that break most automation tools. On top of that, Zalos has developed specialized FinOps libraries for data extraction and harmonization, reconciliation, and other canonical finance tasks, giving it performance that generic automation frameworks cannot match.
The customer stories are concrete. One Bay Area finance manager was losing eight days per month downloading transactions from corporate credit cards and legacy banks, then uploading them into NetSuite and manually categorizing hundreds of line items. After deploying Zalos, a scheduled weekly process now handles 2FA, pulls and submits the required transactions, and categorizes more than 60 percent of outstanding entries automatically. The manager reclaimed two days per week — time redirected into closer inspection of debt covenants, catching cash leakages from overly generous supplier terms, and becoming a more strategic partner to the CEO.
Zalos deliberately runs silently in the background, with no browser window required, and continuously improves itself over time. The company emphasizes enterprise-grade security and records every automation end to end, providing an audit trail for compliance-minded finance and risk leaders. For CFOs, the value proposition is sharp: scale finance operations without adding headcount, reduce human error, and turn a system that once slowed the business down into one that gets smarter with use, propelled by evaluations and reinforcement learning-style fine-tuning.
The timing is especially sharp. Finance leaders are under pressure to do more with lean teams while the surface area of systems they must manage keeps expanding. Computer-using agents represent one of the most promising new categories in applied AI, and Zalos is an early, well-focused entrant with a clear wedge into a buyer that feels the pain every single month.