Sarus software lets data scientists and analysts work on sensitive data without accessing the data directly. They can carry on their work as they always have, from standard database analysis to AI models with complex machine learning training, without risking a leak of any sensitive piece of information.
Traditionally, companies have relied upon data masking, sometimes called de-identification, to protect data privacy. The basic idea is to remove all personally identifiable information from each record. However, a number of high-profile incidents have shown that even supposedly de-identified data can leak consumer privacy.
This marked the beginning of differential privacy which in the simplest form inserted “noise” into the data so the output could not be subject to earlier attacks that leaked consumer privacy. However, by doing so the entire dataset can change, which poses a problem especially with machine learning, where solutions are found by looking for patterns across the data.
Sarus built on this to create a framework for differential privacy, to guarantee that analysis results do not embed personal information and that individual reidentification is impossible. The API provided by Sarus is designed to be completely intuitive and hosted by the client so that the software runs directly on their data infrastructure and no sensitive data leaves their systems.
Technology to Disrupt
At Zillionize we focus on the hard technology space and look out for new ideas which naturally disrupts everything that comes before it. What drew us towards Sarus is their idea for differential privacy as a framework to effortlessly address the risk of privacy breaches at the very core of the problem. Sarus is disrupting everything else that came before it through simplification and automation of privacy protection.
We believe that new companies that possess these criteria can be so game changing that they could spawn a whole new market that doesn’t exist yet. The way Sarus are approaching privacy protection will make traditional anonymization methods currently in use irrelevant, which is truly inspiring to get behind.
Stand Out Founders
No matter how good the idea is, it’s the people who execute on the idea which will make or break it. When we invest in a startup we put emphasis on the qualities of the founders.
Sarus was founded by Maxime Agostini, Nicolas Grislain and Vincent Lepage. All three co-founded AlephD, a ML startup solving publisher monetization challenges which was later acquired by Yahoo/AOL (Verizon) in 2016. On top of that they publish numerous research papers and contribute to open-source software relating to privacy protection.
I discussed about the product with Maxime over a video call and was very impressed by his and the entire team’s knowledge and passion. It was an easy decision to invest in Sarus.
Sarus recently graduated from Y Combinator’s W22 batch and we are excited to have them in our portfolio. They closed an impressive pre-seed round totalling $2 million earlier this year.