As businesses work with data, one of the big hurdles they face is making sure they don't expose personally identifiable information (PII) or other sensitive data. Painstaking manual effort is usually required to get rid of that data. Gretel, an early-stage startup, wants to change that by making the anonymization of datasets faster and easier. Today the company announced a $ 12 million A Series led by Greylock. The company has now raised $ 15.5 million.
Gretel founder and CEO Alex Watson says his company was founded to make it easier to anonymize data and unlock data sets that were previously out of reach due to privacy concerns.
"As a developer, you want to test an idea or create a new feature and it can take weeks to access the data you need. So it essentially boils down to getting approvals to get started, then taking a snapshot of a database and manually remove what appears to be personal data and hope you have it all ",
Watson, who previously worked as a GM at AWS, believed there was a need for a faster and more reliable way to anonymize data, which is why he founded Gretel. The first product is an open source synthetic machine learning library for developers that eliminates personally identifiable information.
"Developers use our open source library, which trains machine learning models on their sensitive data, so while the training is taking place we are applying something called differential privacy, which basically ensures that the model does not store any secret details for the individuals within the data, "he said. The result is a new artificial dataset that is anonymized and safe to share across a business.
The company was founded last year and this year they actually used to develop the open source product and build an open source community around it. "So our approach and our go-to-market here is that we have opened our underlying libraries and will also build a SaaS service that makes it easy to generate synthetic data and anonymous data at scale," he said.
As the founders build the company, they are looking to build a diverse and inclusive organization, something they discuss during their regular founders meetings, especially as they try to take these investment dollars and start hiring more senior people.
"We make a conscious effort to make different candidates apply and to really make sure we reach them and have a conversation, and that has paid off, or is in the process of paying off, I would say, with candidates in our pipeline. right now. So we're excited. It's extremely important to prevent the group from thinking this happens so often, "he said.
The company doesn't have paying customers, but the plan is to build the relationships it has with design partners and start cashing in next year. Sridhar Ramaswamy, Greylock's partner, who is leading the investment, says his company is betting on a pre-billed company because he sees great potential for a service like this.
"We believe that Gretel will democratize secure and controlled access to data for the world in the same way that Github democratized access and control of source code," Ramaswamy said.
