Apostille-USA began with about $2,500 and almost no room for inefficiency. When CEO Rugi Kavamahanga started the company in 2023 with co-founder and CTO Ralph Reijs, hiring ahead of revenue was not an option. Kavamahanga had recently come through the collapse of another startup and was helping care for his father, while the new business was entering a service category where mistakes could delay a visa, relocation, or other major life plan.
Those circumstances pushed the founders to think about growth differently from the start. If the company was going to handle more customers, it needed systems that could absorb more of the work without requiring a new hire every time volume increased. That meant building repeatable processes early and being deliberate about which parts of the operation required human attention.
Apostille-USA helps individuals and organizations authenticate U.S. documents for use abroad, a process that can involve different state, federal, and international requirements.
The model has remained unusually lean as the business has grown. Apostille-USA has now served customers in more than 50 countries and built relationships with more than 100 immigration and global-mobility partners, while cumulative sales surpassed $1 million in 2026. The core team is still just the two founders and one employee.
Building Around the Workflow
Kavamahanga first encountered the problem behind Apostille-USA while preparing documents for a Brazilian digital nomad visa. What seemed like a minor administrative requirement quickly became a process involving unfamiliar authorities, unclear instructions, and little visibility into how the documents were supposed to move.
After learning that immigration professionals regularly saw the same frustration among their own clients, he began researching the market directly and contacting law firms around the world. The challenge was not simply finding demand. It was building a company capable of handling that demand without turning every new customer into more manual work.
Reijs approached that problem through workflow design. His background in software architecture and automation led the company to separate predictable execution from the moments where someone actually has to interpret information or make a judgment.
That distinction became the basis for how Apostille-USA uses AI. It does not try to insert AI into every part of the operation. Traditional automation handles work that follows clear rules, while AI is used more selectively when the information arriving into a workflow is messy or incomplete.
A customer address is a simple example. If the information has been entered incorrectly or in an unusual format, an AI step can interpret it according to the company’s internal rules before the existing workflow continues with shipping and customer notifications. If the uncertainty cannot be resolved confidently, the case returns to a person.
The useful part is not that AI performs the whole task. It is that a small amount of reasoning can be placed exactly where an otherwise automated process would have stopped.
Giving a Small Company More Memory
Small companies often run on what a few people remember. Someone solves an unusual case, learns something important, and may need to reconstruct the same answer months later because the knowledge was never captured in a form the rest of the organization could use.
Apostille-USA has tried to design around that problem by documenting rules and breaking workflows into smaller components. Instead of expecting an AI system to understand the entire business at once, the company can retrieve the instructions relevant to the task in front of it.
For Reijs, that is as much an architecture problem as an AI problem. A larger model or a larger context window does not automatically produce a better system if the information being supplied is poorly organized. Reliability depends on giving the model the right context and keeping its role narrow enough to evaluate.
The company eventually began referring to this broader operating approach as Dynamic Workflow Integration, or DWI. It combines conventional automation, AI-assisted reasoning, documented processes, and human review so that each part of the workflow handles the kind of work it is suited to do.
What Small Teams Can Do Differently
Apostille-USA’s experience points to a different version of the AI story from the one centered on replacing employees.
They did not automate an established workforce away. It started small and used technology to avoid building unnecessary layers of repetitive work as the business grew.
That matters because growth creates complexity long before it creates obvious staffing problems. More customers mean more exceptions, more information moving between systems, and more operational knowledge that can disappear if it remains in one person’s head.
AI can help a small company manage that complexity without pretending that every decision should be automated. Used carefully, it can interpret an exception, surface the relevant rule, or help move a case toward the next defined step while leaving uncertain situations with a person.
Apostille-USA says roughly half of its cumulative revenue has been generated in the past 12 months, while its core team has remained at three people. The significance is less about the headcount itself than about what the company has tried to build around it.
AI has given small businesses access to capabilities that once required much larger operating teams, but the advantage does not come from using more AI everywhere. It comes from knowing where reasoning is actually useful, where automation is enough, and where human judgment still belongs.
Spencer Hulse is the Editorial Director at Grit Daily. He is responsible for overseeing other editors and writers, day-to-day operations, and covering breaking news.
