The Pitch is a Milwaukee Business Journal and Wisconsin Inno series that gives a snapshot of a local startup. The Business Journal doesn’t endorse companies featured in The Pitch, nor is this an invitation to invest. To suggest a startup for possible future features, email tnykiel@bizjournals.com.
After using artificial intelligence (AI) technology in applications ranging from anti-poaching efforts in Africa to 3D helicopter scans of Mt. Everest, Corey Jaskolski noticed a common problem: data was a major bottleneck.
Jaskolski is a Wisconsin native who's also a National Geographic fellow focused on exploration and conservation technology, for which he was named the Rolex National Geographic Explorer of the Year in 2020. He started Delafield-based Synthetaic Inc. in 2019 aiming to create a solution to the AI challenges he encountered in the field.
"In all of those projects, we could never get the AI to work as well as we wanted to," Jaskolski said. "The real challenge always was data."
Traditionally, high-quality AI requires the time-intensive and tedious work of manually collecting, labeling and tagging data to train AI models, Jaskolski said. With Synthetaic, users can train AI in minutes without real, often hard-to-access data, making AI technology accessible to new applications.
"It's effectively a way to generate tons more data than we really have," he said.
This isn't Jaskolski's first foray into entrepreneurship. After earning a master's degree from the Massachusetts Institute of Technology, he started Hydro Technologies, which built sensors and imaging devices for clients including the Department of Defense and NASA. He later sold the company to an oil and gas firm, Jaskolski said.
Synthetaic is in discussions about securing a new corporate headquarters location in downtown Delafield after forgoing its previous office space during the pandemic, Jaskolski said. The company plans to grow its employee count from 19 to around 40, he said.
The company has partnerships with the U.S. Air Force and the University of Michigan's Michigan Medicine, as well as another institution it plans to announce later this month.
The product: AI solutions built on synthetic data, including its Rapid Automatic Image Categorization (RAIC) engine that doesn’t require a priori labeled data, and its MEGAN tool, which generates synthetic data from a smaller existing set of data to boost AI performance.
How it makes money: Partnerships with other technology solution providers and software-as-a-service licenses to Synthetaic's products, Jaskolski said.
Size of the market: The current AI market is around $300 billion, according to Jaskolski. However, he added that "this is the tip of the iceberg," because Synthetaic can also reach market segments that have previously written off AI due to data limitations.
Competition: AI.Reverie Inc., MOSTLY AI Inc.
Competitive advantage: Jaskolski said Synthetaic has a deep technical moat with its ability to "build AI on the fly" with limited data, compared with the usual months-long AI development process.
Business it could disrupt: The data-labeling industry; geospatial intelligence and analytics; security anomaly detection, including anti-poaching applications; and medical testing, according to Jaskolski
Founder and CEO: Corey Jaskolski
Key team members: Principal data scientist Brian Goodwin, director of strategy Eric Franz, director of engineering David Richter
Investors: Lupa Systems, Betaworks Ventures, TitletownTech
Capital raised: $4.5 million
Capital sought: $15 million in upcoming Series A raise
Ideal exit: Synthetaic is focused on developing a technology that democratizes AI and delivers the full impact that AI has been promising for decades, and is not considering exits at this point, Jaskolski said.
