| dc.description.abstract | KYC stands for Know Your Customer or Client which is commonly used for customer or business identification process. It is an indispensable part for anyone in the world of payments, banking sectors,credit & insurance companies and business. It is mandatory for anyone to provide detailed information which validates that they are not involved with corruption, bribery, or money laundering. That in turn allows companies to protect themselves by ensuring that they are doing legal business and with legitimate customers. In traditional KYC procedure, customers need to submit application forms along with the physical supporting documents like Proof of Identity,Address,Birth, Incorporation Certificate, Association Memorandum, etc. to the central institutions. With the help of Digitization this collection and maintenance process can be upgraded. Hence in the future we can replace the entire manual, time consuming, paper based, inefficient and expensive process. Blooming of modern technologies such as APIs, security, analytic, AI and cloud storage are the reason behind emergence of digitization in KYC. In this work, we have tried to build an API with the help of machine learning and Amazon web services, which client can integrate with their KYC application for customer verification process. The API will return whether the user is the one who he or she really claims by matching video of user with valid identification proof. The video that the user has to upload contains 7 different poses such as front, left, right, top, down, tilt left and tilt right. Moreover user also has to smile and blink during recording of video. Here video is recorded with all different poses for security purpose so that we can distinguish that there is actual person in video instead of any bot. If at least 4 poses match with valid identification document then we can considered as user is one that he or she claimed. Here we are using amazon's S3 to store videos and ID proof documents. | |
