“AI cannot be developed in a vacuum.”
1. Tell us about your role in Blue Hexagon?
I am a co-founder and CTO of the company.
2. Can you tell us about your journey into this market?
I have been working in the fields of artificial intelligence and cybersecurity for nearly twenty years. Before co-founding Blue Hexagon, I led a team of people developing AI-based applications for Qualcomm and during the course of my career have been granted 330 patents related to security and artificial intelligence
3. How has IT industry evolved over the past few decades?
Advances in processing power have made possible the development and commercialization of more sophisticated approaches to artificial intelligence, specifically the category of machine learning known as deep learning. But the big breakthrough occurred in 2012 when, at the ImageNet large scale visual recognition challenge (LSVRC), a team of researchers from the University of Toronto, led by Dr. Geoffrey Hinton (“the Godfather of Deep Learning”), introduced an application of deep convolutional neural networks for image recognition. This breakthrough application proved that AI and deep learning were ready to break out from limited applications and be used to solve big problems confronting business. Hinton has since been awarded the Turing Prize for his pioneering work in artificial intelligence.
4. What are some of the major reasons according to you for increase in Cybercrime in recent times?
Cybercriminals are motivated by the prospect of easy money—or fame. State sponsored hackers are motivated to steal secrets or create chaos on behalf of their countries. Both are driven to innovate and have access to sophisticated tools, such as automation, that allow for the generation and unleashing of millions of new malware variants per year. It is relatively easy to get started in cybercrime and there are resources and toolkits available for free or relatively inexpensively. Today you don’t even need to know how to code to become a hacker. This low barrier to entry is a major factor in the increase in volume of cybercrime.
5. How does your real-time deep learning platform protect networks from threats?
Our application of deep learning makes it possible to inspect network traffic—files/attachments and headers etc.—coming into and moving through an organization’s network in real-time to identify threats. We put a lot of effort into training our algorithms to identify threats with a high degree of accuracy (over 99.5%) and in under a second. This includes unique strains of malware known as “zero-day” threats. Our results compare to traditional methods of threat detection that operate in the 12%-20% efficacy range!
6. How has deep learning made cybersecurity more efficient?
The speed (real time) and efficacy (99.5+%) of deep learning represent a transformational shift in the potential for achieving true prevention, which is cybersecurity nirvana. Since we announced our company and platform in February, we have further developed our product to work in cloud native environments and to enable real-time inspection of encrypted network traffic. Further, we’re working with partners to integrate with their existing security tools to extend the advantages of deep learning to protect as many aspects of the modern enterprise as possible.
7. What are some of the tips you can give to our audience to keep their enterprise networks extra secured?
Aside from “buy Blue Hexagon?” According to the latest Verizon Data Breach Investigations Report, 94% of threats to the enterprise are delivered through email, while 23% are delivered via the web (there is overlap as some attacks use email and web in tandem). If you can raise awareness among staff of the risks their online and email habits pose and implement strategies to identify phishing campaigns, for example, you can help reduce the risk of human error. But the fact remains, when you consider that there are four new attacks launched worldwide every second, that is a scale and speed that humans are not capable of matching. We need tools to help us do what we cannot do on our own.
8. What advice would you like to give to the technology Start Ups?
Success in our market not only demands a top-notch team that is dedicated and resourceful, but it requires excellent and committed partners to support every level of the development of the business, including finance, product development and testing, and market entry and expansion. I would also say that deploying early, being customer centric in addressing their unique threat problems will make your product better. AI cannot be developed in a vacuum.
9. What is the Digital innovation according to you that will mark 2019?
We’re biased, but artificial intelligence in all its forms is proving to be a major driver of innovation. Cybersecurity is a small slice of a pie that includes autonomous vehicles, medical devices, medical and pharmaceutical research, telecommunications services and devices, voice recognition services… McKinsey reported last year that AI-based innovation represents a nearly $6 trillion (with a T) global market opportunity.
10. How do you prepare for an AI-Centric world?
When artificial intelligence works best it works invisibly, with no user interaction needed. For example, deep learning is integral to digital photography, identifying subjects and automatically focusing and making adjustments for light. This is analogous to how AI can help people do things they are not capable of doing on their own. Innovators will develop products that meet our needs and we need only go about our business.
11. What are the major developments you are planning, in recent time?
I am not at liberty to be too specific, but we are working with a number of large, well-known consumer, business, and cybersecurity organizations to adapt our technology to extend greater security to every aspect of digital life. Keep your eyes peeled; some of these announcements will happen very soon.
12. Can you tell us about your team and how it supports you?
We launched with a lot of the team that was integral to AI research and development at Qualcomm, and so there is a great deal of experience, trust, and familiarity with the vision and goals we’ve established at Blue Hexagon. In addition, we extended the team with cybersecurity experts from FireEye, Amazon, Symantec, and Palo Alto Networks to round out the capabilities. Having a diversified team in both AI and cybersecurity–and more importantly cross-pollinating ideas between them–has been instrumental to our current success.
13. Which is the Movie that inspires you the most?
Wargames, of course. Many of us in hacking and computing grew up with this movie and it put us on a course to explore what was possible to do with the right technology skills. It was the seminal movie for geeks of my generation. Honorable mention for the movie Sneakers.
14. Can you give us a glance of the applications you use on your phone?
Wag, Yelp, Skype, Plants v/s Zombies, NextDoor, Lots of home automation apps (Alexa, Next, Wink, Arlo)
Saumitra Das is the CTO and co-founder of Blue Hexagon. He has worked on machine learning and cybersecurity for 18 years. As an engineering leader at Qualcomm, he led teams of ML scientists and developers in the development of machine learning based products shipped in hundreds of millions of devices from phones to wearables and IoT. He has also worked on machine learning and security at US-CERT, Intel and Microsoft Research. He is a prolific inventor with 368 worldwide and 163 U.S. granted patents and an additional 400 patents pending. He has published peer-reviewed original research extensively (15 journals, 26 conference, 2 book chapters, IETF networking standards contributions) with more than 2400 citations including a best paper award. He holds an M.S. from Carnegie Mellon and a PhD from Purdue University.
Blue Hexagon is a deep learning innovator focused on protecting organizations from cyberthreats. We've built the industry's first real-time deep learning platform for network threat protection. We have a world class team of deep learning and security experts innovating and setting a new standard for cyber defense. We are headquartered in Sunnyvale, California.