Portland’s Synaptiq started by training AI models but is now helping businesses build AI-driven products, especially those that deploy computer vision — a technology that enables computers to derive information from images.
Long before he founded Synaptiq, a Beaverton company devoted to helping businesses build AI-driven products, Stephen Sklarew was an aspiring biologist studying freshwater ecology in Virginia. He saw himself in a career outdoors, “fishing and hiking and all that kind of stuff. I didn’t want to be stuck behind a desk.”
But then Sklarew got injured and, while recovering, was immobilized in bed for a few months. It was then that his father showed up with a computer and a book titled How to Build Cool Websites.
“Hey, you might as well do something productive while you’re stuck in bed,” Sklarew recalls his father telling him.
And so Sklarew learned to build “cool websites,” beginning with a fly-fishing site for Virginia anglers. People posted fishing reports, weather conditions and other useful information to the site, which was one of the first like it. For Sklarew, seeing his convalescent project make a difference in people’s lives was not only eye-opening but career-shifting. Tech was alluring enough to lead him away from streams and lakes and toward a job at Ernst & Young, the business management consulting firm.
“Before I knew it, I was working in New York City and Boston on big projects, building software systems for companies,” Sklarew says.
Eventually, that would lead to starting his own company, Synaptiq, a Beaverton-based tech firm that got its start by training AI models, then moved on to building data platforms and offering data-strategy services and product development.
But before all that, Sklarew worked for companies big and small, and co-founded a company in Virginia focused on business contact information for marketing departments. The company amassed tons of data, and Sklarew, the chief technology officer at the time, realized that his software engineers didn’t know how to make sense of it all. So they hired Tim Oates, a data scientist who also teaches at the University of Maryland, Baltimore County. Sklarew and Oates worked together for several years until Sklarew decided “it was time for my family and I to get out of the rat race of Northern Virginia.”
Sklarew moved to Oregon, where he worked for Sage software and then for a startup. But about six years ago, he realized he wanted once again to build a company of his own. When a project that involved artificial intelligence came his way, he again reached out to Oates for expertise. By 2017 Synaptiq was making enough money for Sklarew to quit his day job and focus on the company full-time.
“Looking historically in my career, I was graduating college when the web was moving from academia to commercial. And I had started a company when mobile became a really exciting thing,” Sklarew says, “but I was never there before everybody else. I was excited to have one point in my career where I could get to the wave before it hit its crescendo.”
Synaptiq currently has nine full-time remote employees, three business partners and a roster of independent contractors. They billed $1 million in business in 2018 and have grown each year since, with the exception of 2022. They expect to see $2.5 million in revenue in 2023 and have booked work valued at $3 million for the early part of 2024. “Next year we are hoping to double or triple,” Sklarew says.
He credits Synaptiq’s recent growth to the rise of AI awareness, fueled in part by the debut in late 2022 of ChatGPT.
One of the company’s newer clients is a large construction firm that seems an unlikely fit for AI-driven products, Sklarew says. But younger employees are pressing for the use of AI.
“There’s a generation of turnover that’s going to be happening here in the next five to 10 years that will have pretty massive opportunities for digital technology,” Sklarew says.
In the earliest days of Synaptiq, Sklarew focused on reaching out to his network, sussing out who was ready to dip their toes into AI. He and Oates found that, once again, data management was critical. They realized many companies needed help building both the technology to support their data and the human processes to ensure that data is high enough quality to be used for AI.
In 2019 many companies began talking about data as their top asset for identifying new revenue models, Sklarew says. That’s when Synaptiq began offering data-strategy services. They help businesses identify their goals, then help them understand how data supports those goals — and what they need to do with data to expedite getting to those goals.
Sklarew describes Synaptiq’s mission as helping other companies apply AI technologies, often by building products for them that use artificial intelligence to maximize existing data. Their ideal client is not a technology company, he says. It’s a company that wants a partner with expertise in AI.
“What we realized a couple of years in was the hard part isn’t training the models; the hard part is the data,” Sklarew says. “And so it’s not the fun and exciting stuff to talk about, but really helping clients focus on their data such that they can use AI was actually a much larger opportunity.”
Synaptiq has focused on using computer-vision technology to solve problems, an outgrowth of Oates’ research and his connections as a university researcher. Computer vision, one field of AI, enables computers to derive information from visual information, including video and other digital images. Then the programs can make decisions or make recommendations based on the visual information.
In one of Sklarew’s favorite examples, municipal governments and trash-disposal companies can identify how much waste is in a Dumpster based on an image of it, making it easier to more accurately charge tipping fees.
One of the company’s main clients is a large immigration law firm that processes vast amounts of paperwork; computer vision can help analyze that data and automate some decision-making processes. Another legal client is using AI to create a better experience for veterans when they apply for disability claims.
Other clients include: a maker of software for HVAC technicians; a company that deploys low-orbit weather balloons to detect potential wildfires; and a client that needed a tool to help monitor growth in coral-restoration farms.
“They’re not a technology company, but they have all sorts of ideas on how to automate parts of their business using AI, which is a perfect fit for us,” Sklarew says. “They have a lot of data, and we can make a real impact on their business by creating products for them and ultimately for their customers.”
Many of Synaptiq’s applications are in health care. Among their spin-offs in 2019 was Medicine in Motion, a company using AI technology from smart watch data to offer real-time movement prescriptions. They’ve also worked with Spine by Design, a spin-off company that came out of research done by Morgan Giers at Oregon State University. Synaptiq helped create software that uses computer vision and machine learning to analyze patient data to predict the likelihood of reherniation after spinal surgery.
Giers partnered with Charla Triplett, a health care investor who manages Launch Oregon, a University of Oregon innovation lab that helps turn promising academic research into commercially viable ventures. Spine by Design needed support to build something that looks more like commercial software.
Spine by Design hired Sklarew’s team to help the company from the academic version of its software to the next stage, where staff could test it with their clinical hospital partners to “make sure that our model works the same as what we thought it would,” Triplett says.
Working with Synaptiq helped the company get to that validation checkpoint — which is helpful in its fundraising process, Triplett says. Hospitals and insurance companies want products that help predict outcomes, Triplett says. Eventually, they’ll likely sell Spine by Design to a health care software company that can integrate it into existing patient-care systems used by doctors.
For Synaptiq, Spine by Design is an example of how AI is about giving humans superpowers, not taking away their jobs. That’s why Synaptiq’s tagline is “The Humankind of AI,” Sklarew says.
Those sorts of ethical discussions come up all the time in the graduate school class Sklarew teaches to MBA students at Portland State University. AI is merely a mechanism to use data to make decisions, but if humans aren’t reviewing data for bias, Sklarew says, AI is just exacerbating a problem that already exists.
“It’s the hard work,” he says. “It’s getting into the trenches and looking at everything you have on a regular basis and making sure that what you have is representative of the people that are going to be using that data. And if you don’t have good representation of that data, making the effort to collect data that is more representative. That’s hard work, honestly.”
Oh, remember that first “cool” website Sklarew made back in the 1990s as a grad student? He’s come full circle. One of Synaptiq’s side projects is using computer vision (via smartphones) to identify different types of aquatic insects so biologists can better understand stream health.
“That’s something that I used to do behind a microscope for hours at a time as a grad student,” he says. “So connecting the dots there has been great.”
Click here to subscribe to Oregon Business.