We started this crazy venture in 2012, focused on one big, hairy, audacious goal of protecting the world’s endpoints from attack in a completely different way. The same way our own brains work. By leveraging advanced math principles to learn from our past and create the world’s first security machine learning system, we’ve achieved something just short of magical. We've given our system the ability to determine if a file is GOOD or BAD instantly, even if we've never seen the sample before.
When you first hear our claims you might not believe them – you might even say they sound like science fiction – but that’s OK, you're not alone. We've all been burned by security vendors that claim to protect against zero-day threats, only to find out it was all smoke and mirrors. Why wouldn’t you be skeptical?
Rather than try to convince you with talking points, we want to show you how effective our technology is LIVE. We're taking the show on the road to cities around the US to demonstrate our effectiveness versus the competition, and give everyone the chance to prove us wrong. That's how much we believe in our technology.
We're not perfect. Hard stop. No vendor is perfect, but unlike the rest of the security industry we will NOT claim to stop 100% of threats. What we CAN show you is that we stop 99% of "previously undetectable" malware. We believe in our technology so much that we're challenging everyone to bring their worst malware samples to one of our events, in order to see how our tech stacks up them in front of a live audience.
What we've built is something akin to putting a brilliant security researcher in every computer. Think of our product like that 30-year cybersecurity veteran who knows the adversary, their techniques, their tools, and their procedures. They know your defenses and weaknesses, vulnerabilities, and vectors of exploitation. They know the best practices of behavior and computer usage. They know all the tricks and traps. They also know what to look for when it comes to maliciousness.
Let me give you an example. If you receive a file attachment in an email that is an executable file but has a Microsoft Word icon embedded in it to deceive you into thinking it is a MS Word document, then a knowledgable security professional would rightly assume that it's part of a phishing campaign.
Another example, take an executable file that has sections inside it that are set to read, write and execute privilege. That file will be 99%+ bad. Think of a file that has low, low entropy of those same executable sections (like close to zero). That will be 98%+ bad. Now combine all those together and you get a high confidence score that the file is bad and should NOT be executed. Now imagine looking at six million features for every file trying to execute! And you get 99%+ confidence that the file is bad and can block it from running.
Don’t believe we can do what we say? You're not alone. Come prove us wrong.
We'll see you on the road.