KOMAL AGGARWAL

Collaborating and working with DS team for model development and implementation of fraud detection. Optimizing the existing client strategy to provide the optimum results on decisioning. Implemented groovies for 15+ clients & data analysis for 20+ clients. Post Implementation testing was enacted for 30+ clients by collaborating across teams. Collaborated with multiple teams for groovy automation & template creation doubling the work efficiency. Feature Engineering: Evaluating the special cases of fraudulent behavior. Checking the impact of variables on the model. Performing research on the creation of new variables.

Automated and created live dashboards integrating the Biq query data along with visualizations. It consists of the impact of how the business is affected by the campaigns, notifications sent, brands, day type and other attributes. Added visualizations for ease of understanding and better decision taking of business. Performed various campaign level, user level analysis of how business is impacted by user type, platform type. Automated Paytm PVR Contribution in term of tickets to the overall business. So, to understand the market share and position of paytm.

Project work & exposure to risk management: Client interaction, data handling and cleaning. Converting the raw data to the cleaned form by removing outliers, duplicity, and other errors. Using the clean data in CAT modelling to determine the loss numbers and for monthly portfolio analysis. Determining the loss numbers specifically for different perils which ultimately helps in deciding whether insurance should be given or not to that client and if to be given the pricing of the premium to be charged. Also performed data cleansing and exploratory data analysis using matplotlib and seaborn to know relationship between various inputs and the outcome. Developed a logistic model to know which users click on ad. Predicted values for test data and evaluated the model performance.

Project work under ISP Sector (May 2019-June 2019): Studied data available in-service sector to prepare a research report regarding compilation of index for port sector. Compared the data and methods and identified the bottlenecks of the process. Provided variables, methodology that can be used for index compilation of port services. Project work under Economics Census (June 2019-July 2019): Worked on 6th economic census under the economic statistics division. Prepared dash board for the data visualization of inter-state and intra-state comparison of economic census data. Identified the percentage of errors in recording inter-state and intra-state economic census data.