Digital Customer onboarding

Project Title

Digital Customer onboarding

Project Area of Specialization Artificial IntelligenceProject Summary

Digital Customer onboarding system will automate the customer onboarding process of financial institutions. This systems can also be used for other domains as well with little modifications. 

It will allow customer to complete the onboarding process either by capturing live image or by uploading images. The person details will be verified using NADRA database and other blacklisted sites. If person captures live snap and face similarity threshold is greater than or equals to 70%, document similarity index is is greater than 60% and person is not blacklisted than their application is approved. If person upload images of document and selfie and their face similarity index is greater than or equals to 80%, document similarity index is greater than or equals to 60%, and person is not blacklisted than their application is approved. If face similarity index is between 50-70%, and/or document similarity index is between 40-60% , person is not blacklisted and person captures live snap or if face similarity index is between 50-80%, document similarity index is 40-60%, person is not blacklisted, and customer upload images than their application is forwarded to human verifier. If customer is blacklisted  or face similarity index is less than 50% or document similarity index is less than 40% their application is rejected. Human verifier process all the queued applications. 

The system management side interface would include Admin, manager and human verifier. Admin would act as super admin, and can view reports and add management side users as manager and human verifier.

Project Objectives

To automate the customer onboarding process, process which completed in 3-4 days would be completed in 25 minutes

Project Implementation Method

A web application that includes modules of data extraction, document classification, document verification, facial verification, NLP for converting Urdu fields of CNIC into English, and data pipelines.

For data verification, document verification and even for face verification NADRA API is required and it is not available so it's depiction also neede.

Benefits of the Project

This system will reduce the workload of financial institutions by automating their customer onboarding process. 

Technical Details of Final Deliverable

This system would be a web application. OCR wil be used to extract data from document. Document can be a CNIC(smart card or older CNIC) or passport. Data feilds of older version of CNIC would be converted into English using NLP. Face verification will be implemented using deep learning. Data mining will be used to check person's is blacklisted or not using data from different sites. 

Final Deliverable of the Project Software SystemCore Industry FinanceOther Industries Health , Security Core Technology Artificial Intelligence(AI)Other Technologies OthersSustainable Development Goals Decent Work and Economic Growth, Industry, Innovation and InfrastructureRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	79000
laptop	Equipment	1	70000	70000
GPU	Miscellaneous	1	9000	9000