Food safety monitoring tech startup wins first place at Madrona machine learning hackathon – GeekWire
Food safety is a pressing issue. The latest example came last month when an elusive strain of E. coli linked to romaine lettuce sickened 121 people across 25 states and killed one, demonstrating the existing vulnerabilities for how food is screened for safety and quality.
Now a newly-formed group of entrepreneurs wants to use machine learning technology to help keep food free of harmful bacteria and containments before it reaches the dinner table.
Hyper AI took home the first place prize at a machine learning hackathon hosted by TiE Seattle and Madrona Venture Labs, the startup studio housed inside Seattle-based venture capital firm Madrona Venture Group.
The event featured eight teams who came together last month and spent this past weekend creating startup ideas that incorporated the latest machine learning and deep learning technology.
The winning team, Hyper AI, aims to help the food industry with hyper-spectral imaging tech that can detect everything from foreign objects to deadly bacteria. It plans to deploy edge devices on customer premises and do the heavy lifting for image analysis with machine learning in the cloud.
The group, made up of Amazon vets and experienced technologists, explained that existing solutions are either too manual and expensive, or too specialized. It hopes to use machine learning to improve the food scanning technology over time as it learns how to detect more and more contaminants.
Seattle’s top venture capital firm hopes to lure talent from tech giants with new startup accelerator
“They were able to demonstrate why there is increasing awareness of the issue and demand for new, innovative solutions,” said Mike Fridgen, CEO of Madrona Venture Labs who helped judge the pitches. “They had defined their beachhead opportunity, where they would start, through in-depth conversations with potential food processor customers.”
As the first place prize winner, the Hyper AI team will now meet with Madrona Venture Labs with a chance to land a $100,000 investment and participate in the studio’s new accelerator, which just opened up applications for its first cohort.
Accepted startups in the accelerator will use Madrona Venture Labs resources — expertise in company creation, design, engineering, etc.; access to Madrona’s advisor and investor network; and more. It will be housed in Madrona’s new 20,000 square-foot floor that opens later this summer underneath its existing downtown Seattle office.
Madrona Venture Labs held two similar hackathon events in the past and ended up investing in the winning companies: Intellilearn and Eskro (now called Modus). The studio is focusing on supporting “vertical” machine learning and artificial intelligence startups, as explained in this blog post.
The second place team from last weekend’s event was FireWise, which aims predict wildfires before they happen. The third place team, HealthShop, wants to help guide healthcare patients to surgery centers.
(Editor’s note: I was one of the six judges at the event. Others included Madrona Venture Labs CEO Mike Fridgen; Flying Fish Managing Partner Heather Redman; Madrona Ventures Venture Partner and University of Washington professor Dan Weld; Koru CEO Kristen Hamilton; and Microsoft GM Sona Vaish Venkat.)