Instart Logic Announces Helios

May 02, 2017

Instart Logic, a company helping brands around the world make their digital initiatives faster, more reliable, announced Helios which uses artificial intelligence to illuminate and solve challenging digital security issues. Web application attacks are on the rise. A 2012 study found the average company’s cost for every minute of downtime during a DDoS attack was $22,000. Helios, part of Instart Logic’s AppShield family of end-to-end security solutions, uses artificial intelligence, interpretable machine learning, and automation to discover subtle web breaches hidden in billions of applications logs that have been nearly impossible for humans to detect using traditional security technologies.

Helios says it learns from experience, gets better over time, needs no training to get started and provides security experts with fast discovery of anomalies they would not otherwise have found using traditional methods. Most importantly, Helios takes a "no one size fits all" approach, evaluating logs from all incoming site traffic and continuously searching for the “unknown of unknowns.”

In application security, a key challenge is finding anomalies. A traditional system relies on predefined rules to catch attacks. However, previously unknown and subtle attacks cannot be caught by rules. Helios was built expressly to discover previously unknown anomalies and subtle attacks in systems operating at cloud-scale.

The second challenge Helios addresses is the size and scale of today’s log files. Existing security technologies simply cannot keep up with the immense volume of log data. When attacks are found, discovery time can be so long that the attacks can cause great harm by the time they’re discovered. Helios automates and speeds up the attack discovery process by using artificial intelligence and high-performance computing. Helios is capable of analyzing billions of log messages in minutes.

The third challenge is that both web application code and attacks change every day, and new vulnerabilities are constantly emerging. Whatever solution an enterprise uses must accommodate constant change and keep security policies up to date. With so much change, there is minimal room for error. So supervised machine learning algorithms that rely on training datasets fail to solve this problem. Helios uses advanced algorithms that are able to discover anomalies without the need for training.