11.15-16.23-RD-Pyte

SADEGH RIAZI: Hello, everyone. My name is Sadegh Riazi. I'm the co-founder and CEO of Pyte. I started the company together with Dr. Ilya Razenshteyn, who received the Best Thesis Award from MIT in 2017. We created Pyte to make data accessible instantly and globally without compromising data security and privacy. The name of the company reflects that very goal, private byte. We believe that data collaboration should take place over Pytes, not bytes.

Let me start by reviewing three industry sectors that can greatly benefit from AI, but are currently blocked due to concerns revolving around data privacy and security. The first one is drug discovery. Creating a new drug takes approximately eight years and $2 billion. Could Pyte help? Per one study, Pyte can offer time and cost savings of at least 20% to 25%. But necessary data privacy regulations impose a huge hurdle. Even data comparisons within companies take months to review before it goes ahead, creating a major slowdown for AI powered innovation.

Snapshot number two, a whopping 5% of global GDP is being laundered annually. And needless to say that this money is not used in the best way possible. If banks were able to collaborate seamlessly and use AI, they could fight money laundering far more effectively. But financial data is heavily regulated because banks need to keep their transaction data confidential.

And then the last one is around personalized advertisement. I think most of us have felt that our privacy is being violated to some extent when we see a targeted ad. But the truth is that that's a backbone of many online business models. When the new data privacy regulations kick in, access to personal data will be far more limited. The quality and the conversion of ads will go down. And as a result, we the consumers we have to see more ads to compensate for the service providers.

But this doesn't have to be the case. We can create high quality personalized ads while preserving the privacy of consumers. As we all know, without access to high quality data, AI projects are doomed. But our job at Pyte is to make sure that data scientists can accomplish their goals without having access to raw and unprotected data. We have created a software that allows any projects to succeed while fully preserving the privacy of data, not compromising on the accuracy, and using a standard AI frameworks.

These three seemingly contradictory goals can be achieved at the same time thanks to a technology called Secure Multi-party Computation. Let's say there are two companies and they want to understand if Dana is a joint customer or not. What they can do is that they can encrypt their data locally using their own private keys and directly collaborate over encrypted data while keeping their respective customer list private.

In fact, we have shipped our software to Fortune 500 and it's already in use in one case, we helped a health care call center to automatically classify encrypted patients requests without actually seeing the content of the request. In another case, which got started from MIT Startup Exchange Program, we helped a consumer packaged goods company to securely partner with other companies to choose the best scent formula for a variety of their products while keeping their intellectual property fully confidential.

We are looking to partner with any company or organization that has data utilization or data collaboration problem because of privacy restrictions, security concerns, or IP confidentiality roadblocks. In one sentence, Pyte removes the trade off between using data and protecting it Thank you.

[APPLAUSE]