author

Posted by
Dhananjay (DJ) Sampath, on Mar 05 2019

A New Horizon for Enterprise Security

We are at the epoch of a technological paradigm shift.

Historically, with such a shift, there is a tremendous upheaval. This upheaval is always followed by a behavioral change that fundamentally reimagines the way we work, how we play or experience life around us.

For instance, computers changed the way we organized information. iPhones made us explorers as they pioneered hyper-convergence – combining work and play – providing a multimedia experience, where you could click on pictures while reading emails. The applications built on the web reimagined connectivity across the globe, and gave everyone a voice while the cloud fundamentally changed how we grew and scaled businesses.

Companies like IBM, Apple, Facebook, Google and Tesla that identified these new horizons early and executed flawlessly have become incredibly successful.

We here at Armorblox find ourselves staring similarly, at a new horizon. As we build machines that reason with images, with text, or with data of any form, we find ourselves at the cusp of a revolution.


Our Journey So Far

Anand and I caught up in mid-2017 at Zareen's in Palo Alto. We had known each other for over a decade, but that evening, what was supposed to be a 30-minute check-in, turned into a four-hour discussion on how deep learning and natural language understanding (NLU) could make a fundamental difference within enterprises.

Our hypothesis: With connected machines driving increased business productivity and human connections, vast amounts of data – from intellectual property and confidential company data, to sensitive personal data – are constantly being exchanged. Unfortunately, this data is an easy target for attackers. From emails intended to deceive naive users to plain-vanilla data-theft, these attacks cost businesses billions of dollars.

Our solution: If we could build a platform that makes it truly easy to reason about these large volumes of enterprise data, we could move the needle for enterprise security.

Before we started writing a single line of code, we tested our hypothesis with our friends, allies, and customers, and were overwhelmed by consistent feedback that we were on the right track!

We immediately joined forces with Arjun Sambamoorthy and Chetan Anand who had the uncanny ability to see every single step that we'd have to take to build a world-class product.

Together we founded Armorblox.

Our investors Steve Herrod (at General Catalyst) and Ramu Arunachalam (at A-Capital) shared our vision for the future, and fueled the Armorblox rocketship. We are truly grateful for their support and their guidance.

Armorblox – Security Powered by Understanding

Through our conversations with CIOs and CISOs, we learned that more than 90% of enterprise data was textual. The state of the art in enterprise security failed to understand this textual data, and has turned that data into a SHA-256 or a MD5 signature. The conversion into a signature has made it nearly impossible for machines to reason about this data.

At Armorblox, we leverage the latest advances in AI and machine-powered intelligence to apply NLU to improve our defenses against attacks. Using NLU and deep learning, machine comprehension of textual data brings us to new frontiers. Our solution can understand the context of the actual communications that are happening across the enterprise, in emails, documents and other forms of communication.

With every passing day, attacker methods are constantly evolving, becoming more sophisticated. As the cloud infrastructure providers get smarter about security, making it harder to attack, these attackers resort to social engineering to target people! It’s no longer a suspicious email sender domain, a strange attachment, or an unrecognized link. It could be an email from a sender in your address book, in a seemingly normal email exchange.

By modeling human behavior, we automatically create and enforce policies to defend against threats, and rapidly spot abnormalities to accelerate detection. We then leverage the context to reduce remediation workload.

Understanding enterprise data using NLU gives organizations a better way to manage security, protect their customer data, and defend employees against evolving hacker techniques. We’ve built a platform that can reason about textual data at a massive scale, and apply it to solve a variety of use cases – from email security to data loss prevention, to compliance. We’re thrilled to launch out of stealth mode to announce what we’ve been working on. We look forward to working with more companies to solve their biggest challenges.