We work with all traffic sources, both existing and to come. The list of supported ad networks includes: Facebook, Google Ads, TikTok, Bing Ads, Snapchat, Yandex, myTarget, VK, ZeroPark, ExoClick, Taboola, Outbrain, TrafficStars, TrafficJunky, MGID, and hundreds of others.We detect threats in all traffic types regardless of its technical intrinsics: social, mobile, native, push notifications, teasers, popunder, you name it.
VLA™ stands for “Virtual Learning Appliance.” It is the trademark of our state of the art machine learning technology used to detect the most elusive of threats that are actively disguising themselves and manage to get past all of the other filters. VLA detects and blocks new dangers as they emerge, with no human intervention.VLA inspects browser fingerprints and other technical aspects of every visitor, scanning through thousands of features. With the help of computer science, probability theory, and statistics, it finds recurring suspicious patterns in observed facts and yields a certainty of danger, in percents, much like saying, e.g., “I am 97% sure that this fingerprint belongs to someone dangerous, so you better block them.”
HyperLogLog is an algorithm used by our unique traffic filter of the same name. And when we say “unique”, we mean it: this cutting edge filter was invented by Adspect to combat the spreading practice of manual reviews using residential proxies and real devices in target countries. Internals of the HyperLogLog filter are guarded as a trade secret, so there are no analogues of it in existence.Long story short: HyperLogLog allows us to evaluate statistical data from billions of clicks in real time without sacrificing a millisecond of extra overhead. This data contains a lot of distinct patterns that let us tell ad moderators from legitimate visitors. And the more traffic Adspect processes, the more precise this filter will get, i.e. it is getting better and better right now!