Kaivalya Chaporkar

Experience

• 

  Student's Organisation for Technical Activities

  May 2019 - May 2020

  Chief Event Designer
• 

  Atlas Copco India Limited Pune

  Jun 2019 - Jul 2019

  Intern

  In my tenure of internship in Atlas Copco, I worked on project 'Standardization of GA drawings'. In this project I reduced the requirement of separate GA drawings. I successfully reduced the number of GA drawing from 168 to 2

• 

  Principal Global Services

  Jan 2020 - May 2020

  Software Engineer
• 

  Principal Global Services

  Nov 2020 - Jul 2022

  Software Engineer

  • Led business process automation project to resolve false alarms in customer data files, utilizing Java, JS, PL/SQL queries, and data modeling for relational databases by implementing Agile methodology, saving weekly 50 work hours• Utilized Tableau to evaluate code quality reports, identifying areas of improvement, leading to 7% increase in testing code coverage and 5% decrease in code violations• Enabled targeted marketing efforts by accurately predicting customer churn with 96% accuracy for micro business insurance policyholders using a Random Forest Classification model Show less

• 

  PetSmart

  Jan 2023 - May 2023

  Data Scientist

  • Automated customer feedback data collection process by scheduling scripts for scraping 20k customer tweets for PetSmart and its competitors using Twitter API and Tweepy• Conducted sentiment analysis on customer feedback, extracting actionable insights by identifying trending product categories and services as well as successful and failed marketing strategies to improve customer satisfaction• Defined KPIs, created dynamic Power BI dashboard to visualize PetSmart's public outlook and competitor analysis, resulting in cost savings by eliminating the need for outsourcing customer sentiment analysis project Show less

• 

  Circle K

  Jul 2023 - now

  Associate Data Scientist

  • Enhanced sales model performance and interpretability by optimizing feature selection, reducing key drivers from 90 to 10, leading to informed selection of 80% of new store locations dispersing $1.9B.• Deployed revenue forecasting model by employing Neural Nets stacking model that emphasizes site location suitability, boosting performance by 12% with 88% decision accuracy using Streamlit.• Engineered automated end-to-end Databricks workflow to pull data for potential sites, generate its similarity scores against existing stores, and feed to Power BI dashboard nightly manner, cutting down 15 work hours per week.• Conducted training camp for the business team to ensure effective adoption of delivered tool, simplifying technical concepts and empowering non-technical stakeholders to utilize solution confidently and independently.• Implemented KMeans and spectral density clustering to group Circle K stores, transitioning Show less