Helped an Award-Winning New York based cybersecurity firm with
Computer Vision project to create deep learning models by analysing websites and webpages
About the Customer
The Award-Winning New York based cybersecurity firm has its pedigree rooted in top accelerators, well-known cybersecurity investors and former Intelligence Community members. This anti-phishing organisation focuses on cybersecurity and endpoint security. Through computer vision, and deep learning they help companies, businesses, and organizations on website, network, and computer security.
Despite numerous cybersecurity efforts, phishing attacks have been on the rise with 93 percent of all breaches beginning with phishing emails. The client required the assistance of Computer Vision wizards to build machine learning models that were accurate and can scale. Machine learning models needed to be trained on precisely identifying a potential phishing link as soon as it was opened on browsers. Achieving this accuracy called for processing and annotating images, text, and videos of several million web pages.
The client took assistance from NextWealth and their expertise in Computer Vision to create deep learning models by analysing websites and webpages. NextWealth used object detection and bounding box annotation to train machine learning models to identify and label logos of brands to detect if a logo on a webpage or website was authentic or false. Model Validation was also carried out to take certain predefined steps in the event of an attack and to identify false positives and false negatives.
Using deep-learning computer-vision and validated machine learning models, the company was able to build a workflow tool and improve the efficiency of the tool 7X times. They can now detect phishing attacks in real time at the point of click within browsers. The validated machine learning models are trained to identify phishing websites, thus helping clients from major cyberattacks and the costs associated with them, which can amount to millions of dollars and loss of customer trust.