Rishab Panyam Email & Phone Number

Chicago, Illinois, United States

• Developed Custom XGBoost Risk Models to predict the likelihood of default of a loan. The Custom Models outperformed the off-the-shelf scores by 20%. Used Reject Inference techniques to get an unbiased population.
• Automated Quarterly Market Overview Reports using Spark,R and Excel Macros reducing the time for the Report generation from 4 weeks to 1 week leading to a 400 % reduction in report generation time.
• Developed Retrospective and Risk Tier Reports for clients to revise their decisioning process and select the best scores for decisioning. Used Feature Importance Techniques to find top n features to use for initial.

Pittsburgh, Pennsylvania, United States

Evaluated the performance of multiple Machine Learning models such as XGboost,Random Forest,Linear Regression,Survival Analysis in Python to predict the completion time of a batch of warehouse picks.Selected Random Forest Regressor for its performance(a Mean Absolute Percentage error of 10%) and its robustness to overfitting.Deployed the Selected Machine.

Chicago, Illinois, United States

• Performed Feature Engineering and Feature Selection to develop and identify key variables that impact price sensitivity.
• Developed a Price Recommendation model using Python, Sklearn, Numpy and Pandas based on Customer and Item Attributes to recommended prices along with their 95% confidence interval. Obtained a Test RMSE of 0.11 for the.
• Visualized and Presented price recommendations for 33 million unique Combinations of Customer and Items on Power Bi for the business team at LabelMaster.

San Francisco Bay Area

• Performed Regression Analysis and A/B Testing Experiments to track the impact of new features on KPIs.Suggested insights which reduced customer churn by 10%.
• Developed Dashboards using SQL and Excel tracking monthly customer churn and conversion. Created ad-hoc statistical analysis to determine the channels leading to customer conversion.
• Created a web scraper using Selenium, Python and SQL to extract and organize valuable data to improve the product.

Bengaluru Area, India

• Developed an Algorithm for Object Detection of Vehicles using Histogram of Oriented gradients as Features and SVM classifier.
• Improved the IOU accuracy of the Algorithm using Dynamic Window Size to account for change in car dimensions across the image.
• The algorithm was implemented in Python using OpenCv,Skimage,Numpy,Scikit-learn, Pandas libraries.

Mumbai Area, India

• Worked on solving business problems for a top fortune 500 company by deep diving into Nielsen data.Identified the potential reasons for change in KPi’s of the business by performing root cause analysis.
• Developed Spotfire Dashboards to create interactive Data visualization to track KPIs of the products.Developed Back end of the Spotfire Dashboards using Knime and SQL.Improved the efficiency by reducing the time take.

Master'S Degree, Data Science, 3.84/4.00

Activities and Societies: Recipient of the Data Science Excellence Scholarship•Recipient of Data Science Excellence Scholarship. Relevant.

Bachelor Of Technology - Btech, Metallurgical Engineering