Our crawlers are capturing millions of images from 10,000+ websites from over 20 countries daily. Data will be synchronized from 100+ crawlers all over the world to our servers. All images will be filtered and non-ad information will be removed by filtering module automatically. Remaining data will be classified and sent to content team for process. After data validation, new advertisers and new crawing rules will be updated by learning module. All crawlers will be updated with new crawling rules immediately. Besides, a robot will explore new publishers and inform content team. Once it is verified, new publisher will be added to crawling list and being monitored for performance.