It is widely understood that once machine learning models are deployed in production, the accuracy of the results degrades over time. Artur.i was launched in 2019 to help companies track their models to ensure they are aligned with their goals. Since then, the company has added clarification and bias reduction to its array of services.
The device is resonating in the market, and today the startup announced a $42 million Series B round. Company founder Adam Wenchel told TechCrunch that it’s the biggest round of machine learning tracking startups.
Authenticity means protection from bias, and that’s what the company has been working on since we last spoke to it, when it took a $15 million Series A round.
“We’ve done a lot on object bias. It is becoming too dominant for people’s minds, how can they not be biased like these models? And we’ve done a lot of new IP development around how to automatically adjust the results of these models to meet the equity constraints that customers want to achieve.
Explaining, as the name suggests, is understanding why you got the results you did. Wenchel uses the example of high blood pressure, which can be a dietary or non-controllable factor, or it can be a genetic factor beyond your control and may require medication to bring it down. It’s important to understand that there is no one-size-fits-all answer, and it can help prevent overgeneralizing what the machine learning model tells you.
He said he definitely noticed a difference in raising this year compared to last time. As opposed to the bubble environment of 2020 where we had people calling every five minutes, we had to meet with dozens of investors to get those multi-word papers, ‘Are you ready? are you ready? Are you ready yet?’ But everything went well for us,” he said.
Perhaps the company’s growth is one of the reasons for investors’ interest. The startup has averaged 58% ARR growth over the past four quarters, which looks even better when you consider the economic upheavals we’ve faced over the past two years.
The company has 55 employees today, up from 17 when CERA was launched, and Wenchel said diversity remains a company goal, both at the capital desk level and at the employee level.
It’s especially important in the research environment, where having a diverse workforce helps prevent bias from seeping into their software, he said. “We’ve published several papers and the group in particular is incredibly diverse, and I think it’s a better group for it,” he said.
Today’s round was led by Akrow Capital and Greycroft. The Cape Table includes operators Theresa Gow from Acrew and Ashley Mayer from Cooperative. Gow will join the board on a cash basis.
