Analyzed large corporation data for their revenue & cost Performed Data discovery, data value articulation and statistical modeling for revenue, cost & CCI in GCP environmentIntegrated Explainability, Scenario Planning & Predictions, from prophet models, to Power BI dashboard for executives to explore future scenarios

- Built use cases based POCs for a large telecommunication company to effectively address customers calls/experience using predictive modeling like sentiment analysis, regression, rules based engines etc.- Was part of a large Oil & Gas industry's data science practice to transform their digital talent discovery experience and build analytical dashboard for executives to gauge effectiveness of digital learning program.- Was involved with a large fortune 50 retail client in their digital transformation. Worked with their data science practice in evaluating and fixing models affected by closure of large number of stores and unavailability of data in pandemic months. - Managed Care Health Analytics: Analyzed a large managed health care business, developed tableau dashboards and forecasting models in R. Did benchmarking against Fair health, CMS and other healthcare providers data.- Claims Analytics: Built predictive models as part of a bigger initiative to optimize claims processing at a large insurance company using python and machine learning techniques.- Fraud Analytics: Investigated frauds happening, in auto and life insurance, at a large insurance company. Completed Data discovery and presented that to client to give an idea of scale of frauds and how Accenture AI can help mitigate the problems.

Trying to understand life insurance industry, it's jargon and helping Kemper's field agents to focus on lapsing policies using predictive models. Built two predictive models using machine learning techniques i.e. Gradient boosting and Extreme Gradient boosting, for new business and renew business. Renew business has highly imbalanced data and hence XGBoost. Models have been deployed to production and we are generating predicted values every month for to be lapsing polices and giving to field agents to take action. Mapped marketing data to kemper's target population and identified opportunities in unexpolited market based on zip codes.

Health care company from provider sides-----------------------------------------------------•I am involved with laying down the infrastructure for predictive analytics. Python, SAS, and R are being chosen as primary tools for model building. Implemented application of agile methodologies for implementation of data science across enterprise. Interact with IT team to achieve various targets.•Created infrastructure for Risk models, 349 model groups in numbers, using various machine learning algorithms using python and SAS. Risk models are based on patient’s electronic medical records (EMR) which uses ICD9/10 codes along with other variables. These models predict mortality rate of a patient, length of stay (LOS) and total cost a patient will incur during hospital stay based on several diagnosis and procedural factors.•Created an analytic product, in python, which will take 349 model groups’ data sets as input and using various data preparation techniques and machine learning techniques i.e. Random forest, Randomized LASSO, Extra tree etc. perform feature selection to help identify important clinically relevant features for risk models to be built. Result: Huge saving of time and resources in finding important features in existing risk models as well as in case of any update of data sets in future. •Analyzed lab data using various exploratory analysis techniques and explored the impact of incorporating that in existing risk models which were built using regression and logistic regression techniques.•Created hospital profiler where any member hospital can come and compare itself with the peer group on various parameters. Various unsupervised learning techniques related to clustering e.g. k-means, hierarchical and principal component analysis (PCA) were used in this.•Created a time series analysis product, using R, which analyses a huge dataset comprising of more than 1000 SKUs and forecast based on spend.

UHC is a member owned healthcare organization which provides services to almost all academic medical centers of USA. In a similar role, VHA provides services to community hospitals. Together they provide services to almost 35% hospitals across United States. I joined as a lead data scientist to lay down infrastructure and start work for predictive modeling. Organization is getting real time encounter level data and we are exploring various options to provide services to member hospitals using predictive modeling.

I did various jobs related to improvement of processes using different analytics techniques. Worked with claims division to detect frauds, claims error etc. by building various predictive models.

• Worked with actuarial/analytics team to design & develop rating engine, in Java, to produce health insurance rates based on several input criteria/data. Provided subject matter expertise for development of application that provides premium rates in real time to potential customers to comply with Affordable Care mandate for connection to health exchange.• Recognized problem and designed system to reconcile claims data, pharmacy vendor’s data, and behavioral health vendor’s data and ensured that all three remain in sync with each other 24 X 7.

Technical lead of a pilot team of 6 software developers, analysts to develop infrastructure, from the ground, for service oriented architecture (SOA). My job was to build enterprise wide framework for SOA based services development across all verticals of Blue Cross Blue Shield of Michigan. Created in-house 33 initial services and later on when building of services were outsourced, I managed federated build of multiple vendors as services architect.

Worked as a senior software developer for development of retail banking software product of TSYS. Latest Java and wireless based technologies had been used for this product. This product was used by many european banks like Barkley, ABN AMRO etc.

Developed Enterprise's rental car's retail software. It was using the latest Java/Struts technologies.