RAHUL ROSHAN

Previously Worked in Entities Extraction project. The project scope is to learn document dataset and create model to extract key information from submitted documents for Automobile. Created PoC Prototype using Azure Form Recognizer. Creating Web API calls thru Python. Consume data thru Python and persist the data in Azure SQL DBA. Operationalized the model using Azure Functions / ML Ops Concepts. Advanced Analytics PoC for Insurance BPS. Involved in several Data Insights initiatives for Insurance BPS to showcase the value of Data Insights, Prediction / Classification Models. The business process involves in claim processing, document scanning, etc. Working with Insurance BPS Sales team to identify possible business use case and do Data Analysis, building dashboards for prespective / descriptive analytics. Involved in client presentation on the PoCs to showcase the outcomes and work on getting larger operational projects.

Currently Working in an Entities Extraction project. The project scope is to learn the entities from the e-mail body and attachments and create model to extract key information from submitted e-mails for Mosaic insurance. Exploratory Data Analysis(Uni-grams,Bi-grams,Tri-grams,Wordcloud). Created Python Scripts to read data from email attachments(PDF ,docx). Used the existing CRF model for modelling. Analyse the model results to understand and increase the accuracy of Entities. Post-processing of Entities to improve Accuracy. Perform Google search for entities not available in e-mail body.

Tools Used: AWS , Python, Power BI, Azure Data Lake