With around ten customers today, Stacks aims to scale their product to hundreds in the nearest future. "It would be impossible to envision such a trajectory with a traditional infrastructure. Thanks to Google Cloud, we benefit from natural elasticity and automated scaling, which allows us to grow easily," he explains.

Beyond the technology, it was also the quality of support that convinced Stacks to adopt Google Cloud. "Our team has never been treated as a 'small customer' but as a true innovation partner," insists Ivan Bovyrin. "Whenever we encountered a difficulty, whether it was adjusting quotas or technical issues, we could always count on a quick and effective response."

Through the Google for Startups Cloud program, Stacks also benefited from direct access to the resources needed to integrate the latest AI innovations, optimize its infrastructure as it grows, and strengthen the security of its platform.

These AI-powered features significantly reduce manual processing and standardize operations that previously often required specific developments. "AI helps us industrialize tasks that were previously very context-dependent. This allows us to onboard new customers faster, regardless of their data formats or ERP environment, and offer them standardized automated workflows without having to develop specific integrations for each case," emphasizes Ivan Bovyrin.

Today, Stacks continues to enhance its platform while expanding its customer base. The next step is to develop workflows fully driven by AI. This new approach is part of the rise of agents, systems capable of chaining multiple actions autonomously based on a defined objective.

At Stacks, the ambition is to entrust certain well-defined accounting sequences to these agents, based on the company's internal rules and the history of previous processing. AI no longer intervenes only at the task level but across an entire process, from initiation to validation. "We are seeing the emergence of use cases where human intervention can be reduced to zero, provided we have enough context and a quality history," explains Ivan Bovyrin. "The end-of-month reconciliation, for example, often follows identical rules from one month to the next. This type of task can be entrusted to an AI agent that knows when to launch the process, what data to process, and how to validate it. Our goal is clear: to make the financial close an almost instantaneous process, as simple as clicking a button."

Thanks to artificial intelligence, a resilient and scalable infrastructure, and a sustained pace of innovation, the young company intends to transform a historically cumbersome and time-consuming function into a lever for agile and automated management. An ambitious bet — but already well underway.