Retention and Purge (R&P) - This is the ongoing project on which I am currently working on. In this, we are trying to build a framework for R&P where we are Extracting data from multiple source systems, Store it in R&P Repository, execute rules to Identify data which can be purged and send Notification to business systems to purge the specific data which was asked to purge. In return business system sends us an acknowledgement about the purge status. We are doing this to keep the systems compliant as per the new Data Protection policies in US (SOC2, California Regulation)

Responsibilities

● Involved in Designing the whole framework with the help of architects

● Initially Performed the POCs to prove that it is feasible and demoed to Compliance Team

● Working on designing the interactions with multiple verticals e.g. via API, File FTP etc.

● Technology Stack – FLASK, Python, Oracle (cx_oracle), LINUX, Bamboo (CI/CD)

o AWS is an area of exploration we considered here, POC is in progress

We were asked to perform a Sentiment analysis of the customer mails regarding their queries for Claims and Policy Admin data. It will classify the mails into Complaint or Query.

Responsibilities

● Read the customer mails from a JSON file (given)

● Performed Tokenization and removal of stop words

● Used Stemming and TF-IDF to create the Term Matrix

● Applied Classification Machine Learning on top of that to classify into Complain and Query

● Technology Stack – Python, Sklearn, NLTK, Excel

Claim Examiner – We have claim application with auto adjudication feature for the claims. But there are always some claims which needs manual intervention. I was part of developing a system to help the medical reviewers to have the claims classified with confidence score using Machine Learning model built in Python so that they can Approve/Reject them

Responsibilities

● Involved in data gathering from business people and medical reviewer

● Explored and cleanse the data using Pandas Library

● Did Feature engineering to extract, transform and dimensionality reduction

● Technology Stack –Python ML libraries like sklearn, pandas, numpy etc.