A Madison-based startup called Strudel has won this year’s Pressure Chamber pitch competition with its AI-based tech support platform.
The contest was held yesterday afternoon at The Sylvee music venue in Madison as part of Forward Fest, an annual event series focused on entrepreneurship and technology.
Strudel’s co-founder and CEO, Kristin Isaac, beat out four other competitors for the grand prize after each gave rapid-fire, 5-minute pitches and answered questions from a panel of investor judges. She was presented with the “golden suitcase” award, landing her a spot on the Greater Madison Chamber of Commerce’s trip to Silicon Valley this fall to meet with venture capital firms.
“I’m feeling very grateful right now,” Isaac said after accepting the award, applauding her fellow co-founders as well as her mentors and company advisors. “Thanks to all of you, this is so amazing.”
The startup has created a platform to handle complex work that’s often done by software engineers, aimed at resolving technical issues more quickly while saving time for engineering teams. Isaac said software professionals are spending 40% of their time on technical customer support, and Strudel’s AI agent “bridges support and engineering” using various language models.
“We are building the next Epic-sized software company out of Wisconsin,” Isaac said.
She says the company is seeing “incredible traction” after launching earlier this year, with its first three business-to-business pilot customers converting to the paid model and other customer acquisitions underway. The company’s pricing structure “shows a ton of ROI” for its customers, according to Isaac.
“They’re saving tens of millions of dollars a year with our solution, so spending $250,000 or more on our solution is a no-brainer,” she said.
Isaac fielded judges’ questions about how competitors are addressing similar issues, as well as how the company’s AI is being trained. She noted Strudel can keep training data from customers “boxed in” to address potential security concerns during that process.
“We are using [large language models] as well, but our small language models are what allow us to use very specific data sets, which allow us to get super accurate and also operate at a much lower cost than only the LLM,” she said.
Strudel is currently trying to raise $500,000 with $300,000 already committed.
