Zalos AI Ltd. said today it has raised $3.6 million in seed-stage funding to build artificial intelligence agents to automate finance operations.

The round was led by 14 Peaks with participation from Cohen Circle, 20VC and notable angels, including FedEx Corp. Chief Financial Officer Mike Lenz, Happy Robot Inc. founder Pablo Palafox and Indeed Inc. founder Paul Forster, among others.

Modern finance is led by teams that work manually through enterprise resource planning, customer relationship management, spreadsheets, email and banking platforms that don’t have the backend to talk to each other. The “connective tissue” between these systems is often human, manually stitched together by finance operations between billing cycles, bookkeeping and reporting.

Zalos said it built an AI solution that helps ease this frantic system by thinking the way humans do, by understanding deep business context and accessing the same systems that business finance teams do. With the rise of generative AI technology that can ingest spreadsheets, billing receipts, emails, text messages and other visual data, it’s possible to log into separate systems and centralize work that has traditionally been split across the workplace.

“Finance teams have the systems, but they are still doing the work manually because the stack is not connected,” said co-founder and Chief Executive William Fairbairn. “We built Zalos on the belief that CFOs should not need to rip out their existing stack to adopt the latest in AI. We want to start by sitting on top of what is already there.”

The company’s platform can convert screen recordings of finance workflows into computer agents that can log in, navigate screens, enter data and validate data within ERP systems. These are repetitive, time-wasting tasks such as data entry involving password-protected portals involving outdated interfaces, websites, documents, spreadsheets and other documents that need to move data. Moving and validating this data takes time and energy and it’s also error-prone and requires constant oversight.

The company boasts that its system is capable of building not just one agent to do this work, but hundreds — all of them working around the clock to complete tasks.

As agents work, they can check in with their human counterparts, providing historical audit trails for human-in-the-loop management. For private industries, such as finance, stakeholders are expected to hand over reports to regulators.

“Legacy ERPs’ speed of innovation has stalled, leading to growing manual work in the place of transformative automation. AI-native ERPs may offer a credible alternative for companies that have not yet committed to a system,” said co-founder and Chief Technology Officer Zalos Hung Hoang.

AI agents offer a potential clean path to provide automation that sits atop already existing financial infrastructure. Companies do not need to change anything they already do. Zalos’ system understands what a company does, works with it just like a human would, and if it runs into something ambiguous, it checks in.

Once a recording is set or instructions are written, it’s automated forever. It never takes a holiday.

Zalos currently targets midmarket enterprise resource planning, where it already has customers, but with this financing, the company intends to expand its operations beyond that market into enterprise ERPs and on-premise systems.

Through building a wide-reaching context graph for the finance stack and using swarm agents, the company said, it intends to bring its technology to more CFOs and change how finance teams operate.

Image: SiliconANGLE/Microsoft Designer