My Role was to design the architecture of the solution in addition to develop the middleware component. Also, I have developed the Face-Recognition and Document classification applications. It was a three-layered architecture with front-end developed with HTML, CSS, Javascript; Middleware built using python flask application and backend contains Raspberry Pi device. These layers have been integrated using REST API.
For Biometric Verification, we have enabled Fingerprint Verification, Document Classification (Government ID) and Face Recognition.

We have built a Machine Learning model to determine the complexity of the role based on the wordings used in the work experience of an employee. Our ML model infers the complexity using Natural Language Processing. The CV description having unstructured text data is given as an input to the algorithm and using machine learning models, it infers the complexity of the role as High, Medium, or Low.
I have built the pipeline of this ML model.

Client has an old file system based on older version of Document Management(WebCenter Content).We have migrated this older file system to newer document management system.
This Document Management system was integrated with Email Client(Outlook),Microsoft Office for seamless document ingestion. We have built a Java module which would help to integrate the above applications using REST API.
I have led the team which provided support and enhancements to the technical issues pertaining to document management raised by the business users.

I have led the team to build the backend service of a J2EE based portal. This backend service has also been used to expose the different functionalities of underlying document management service. Additionally, helped team to integrate this backend service to Service Oriented Applications using SOAP based webservices. Authentication and Authorization have been implemented using Identity Access Management of Oracle.

We have built a multiclass classification model to categorize articles to different classes.Using tagging ,this model was further enhanced to discover the correct document among a set of documents.

We have built a model which is able to compare two or more documents and ascertain plagiarism with a certain degree of accuracy.This has been used to help check fradulent activities in curriculam vitae.