Long before artificial intelligence became a fixture of business conversation, Joel Yi was using it to identify rare plants. In 2018, years ahead of the current wave of interest in the technology, Joel Yi built a machine learning model designed to recognize uncommon plant species with strong accuracy. That early project, modest as it sounds, helped shape the way the founder of DeployAIBots thinks about artificial intelligence to this day.
The model was an exercise in a specific kind of problem. Identifying rare species requires a system that can learn subtle patterns from limited examples and apply them reliably. For Joel Yi, who studied computer science and earned recognition for his work in artificial intelligence at Pacific Lutheran University, the project was an early demonstration of what AI could do when applied to a concrete task.
It was not a thought experiment. It produced a working result, and that experience left a lasting impression.
What Joel Yi took from it was less about plants and more about leverage. He has described arriving at a realization early in his exposure to artificial intelligence: that one person equipped with the right systems could potentially produce the output of a larger team.
Building a model that could perform a task that might otherwise require specialized human expertise made that idea concrete. The technology was not just interesting. It was a way to multiply what a single person could accomplish.
That insight redirected his thinking about business. Instead of asking how to add more people to handle more work, Joel Yi began asking how to reduce the dependence on people for repetitive tasks.
The plant identification model was an early glimpse of how that might be possible. If a system could learn to do something difficult and do it consistently, then the same principle could be applied to the routine work that consumes much of a company’s time. Years later, that principle became part of the foundation of DeployAIBots.
The early project also instilled a bias toward building rather than theorizing. Joel Yi did not simply read about machine learning. He built a model that worked.
That hands-on experience informs his frequent criticism of an artificial intelligence market he sees as overloaded with talk and short on delivery. He has argued that many people in the field sell ideas and offer advice without ever deploying working systems. His own history began with deployment, with a model that actually performed, and that origin shapes the standard he holds others to.
There is a thread connecting the 2018 model to the systems DeployAIBots builds now. Both are about getting artificial intelligence to perform a defined task reliably.
The company installs agentic AI designed to execute operational work such as scheduling, customer communication, and internal coordination. Its automation is meant to run a process from start to finish rather than merely assist. The ambition has grown, but the underlying belief is the same one the plant model first confirmed for Joel Yi: that AI delivers value when it is built to do real work well.
Joel Yi’s path since then has added other dimensions. After his early academic work, he became one of the first cyber officers in the United States Army cyber branch, where he learned to build and evaluate systems under demanding conditions.
That experience layered a concern for reliability and security on top of the leverage mindset the plant model first sparked. Together, they produced a founder who is both interested in what artificial intelligence can do and disciplined about how it should be built.
The story of the plant identification model also speaks to timing. Joel Yi was working seriously with machine learning before it became a mainstream business focus, which gives him a longer view than many who arrived during the recent surge of interest.
He has watched the technology move from a niche pursuit to a mainstream obsession, and that perspective informs his skepticism of hype as well as his confidence in the technology’s practical value.
For Joel Yi, the early model remains a useful reference point. It was the moment artificial intelligence stopped being an idea and became a tool in his hands.
Much of what he has built since, including DeployAIBots, traces back to that first proof that AI, when applied carefully, can work in practical ways.
