Nayeem Islam He spent almost 11 years with chipmaker Qualcomm, where he founded his Silicon Valley-based research and development center, recruited his entire team, and oversaw research on all aspects of security, including the application of machine learning to devices. mobile and network to detect threats ahead of time.
Islam was nothing if it was not prolific, developing a system for machine learning on devices to detect malware, libraries to optimize deep learning algorithms on mobile devices, and systems for parallel computing on mobile devices, among other things.
In fact, because of his work, he also saw a great opportunity to better protect companies from cyber threats through deep neural networks that can process every raw byte within a file without ignoring anything, and that can uncover complex relationships. within data sets. So two years ago Islam and Saumitra Das, a former Qualcomm engineer with 330 patents to his name and another 450 pending, went off on their own to create Blue Hexagon, a company that now has 30 employees in Sunnyvale, California. . revealing that he has raised $ 31 million in funding from Benchmark and Altimeter.
Funding comes about a year after Benchmark. He quietly led a $ 6 million Series A round for the firm.
So what are investors so optimistic about the company’s prospects, aside from its reputable founders? In a word, speed, apparently. According to Islam, blue hexagon he has created a real-time cybersecurity platform that he claims can detect known and unknown threats on the first encounter and then block them in “seconds seconds” so that malware doesn’t have time to spread.
The industry has to move to real-time detection, he explains, explaining that four unique and unique malware samples are released every second, and he argues that traditional security methods can’t keep up. Sandboxes, for example, which refer to restricted environments that quarantine cyber threats and prevent confidential file breaches, he says, are no longer state of the art. The same goes for signatures, which are mathematical techniques used to validate the authenticity and integrity of a message, software or digital document, but they are being avoided through new, rapidly evolving malware.
Only time will tell if Blue Hexagon is much better able to identify and stop attackers, as Islam insists that is the case. It is not the only startup that applies deep learning to cybersecurity, although it is undoubtedly one of the first.
Critics, some of whom are protecting their own corporate interests, also worry that hackers could thwart security algorithms by targeting the warning flags they seek.
Still, with its technology, equipment and launch, Blue Hexagon is beginning to persuade not just top investors of its merits, but a growing and broad customer base, Islam says. “We all have this problem, from large banks, insurance companies, state and local governments. Nowhere do you find someone who doesn’t need to be protected. “
The Blue Hex can even help clients who are already under attack, Islam says, even if it isn’t ideal. “Our goal is to catch an attack as soon as possible in the chain of deaths. But if someone is already being attacked, we will see that activity, we will identify it and we can turn it off. “
Some damage may be already done, of course. It’s another reason to plan ahead, he says. “With automated attacks, you need automated techniques.” Deep learning, he insists, “is a way to level the playing field against attackers.”