Kavya Basavalingappa Umashankar

Experience

• 

  Verzeo

  Jan 2020 - Feb 2020

  Machine Leaning Intern

  As a Machine Learning Engineer Intern at Verzeo, I preprocessed and transformedcomplex raw data using Python libraries, cutting data cleaning time by 50% and optimizing analysisworkflows. I implemented optimized machine learning models, such as SVM and Logistic Regression,to detect credit card fraud, achieving a 90% detection rate and reducing false positives by 30%.Additionally, I enhanced model accuracy by 10% and reduced false negatives by 15% throughhyperparameter tuning.Achievements:- Cut data cleaning time by 50% through effective preprocessing and transformation of complex rawdata.- Achieved a 90% detection rate for credit card fraud through the implementation of optimizedmachine learning models.- Enhanced model accuracy by 10% and reduced false negatives by 15% through hyperparametertuning. Show less

• 

  UXL Technologies

  Jul 2020 - Sept 2020

  Intern

  As a Backend Developer Intern at UXL Technologies, I developed codes using Python and Java, to enhance the easier workflow of the project. Provided one-on-one tutoring sessions for 2-3 children each week, resulting in a 100% success rate for students' coding projects. Collaborated with cross-functional teams to design and implement new features, resulting in a significant decrease in loading times (by up to 50%) and an overall improvement of platform performance by more than 75%.

• 

  MOVIEQU

  Nov 2020 - Dec 2020

  Data Intern

  During my internship at Moviequ, I developed an automated web scraper using Python’sSelenium, which extracted data from 20 movie websites, reducing manual data collection by 2 hoursdaily. I also designed and optimized SQL scripts for data loading into relational databases, whichreduced processing time by 25% and ensured consistent data integrity. Additionally, I built andmaintained dynamic Tableau dashboards that uncovered 15 critical features, improving userexperience and product decisions.Achievements:- Reduced manual data collection time by 2 hours daily through the development of an automatedweb scraper.- Achieved a 25% reduction in processing time by optimizing SQL scripts for data loading.- Improved user experience and product decisions by uncovering 15 critical features through dynamicTableau dashboards. Show less

• 

  New Jersey Institute of Technology

  Aug 2022 - Dec 2022

  As a Graduate Research Assistant, I led the development of a Python-based machinelearning pipeline for solar flare analysis, improving model accuracy by 15% through SMOTE andparameter optimization. I designed and deployed machine learning and deep learning models,maintaining an average accuracy of 80% across multiple projects. Additionally, I documented datastructures and workflows, implementing CI/CD pipelines with GitHub and Zenodo, which optimizedthe ML lifecycle and boosted performance by 15%.Achievements:- Improved model accuracy by 15% through the development of a Python-based ML pipeline.- Maintained an average accuracy of 80% across multiple machine learning and deep learningprojects.- Optimized the ML lifecycle and boosted performance by 15% through effective documentation andCI/CD implementation. Show less

  • Undergraduate Teaching Assistant

    Sept 2022 - Dec 2022

  • Graduate Teaching Assistant

    Sept 2022 - Dec 2022

  • Graduate Research Assistant

    Aug 2022 - Dec 2022
• 

  CloudData Technology LLC

  Jun 2023 - Aug 2024

  Data Scientist / Data Analyst

  In my role as a Data Scientist at CloudData Technology LLC, I leverage MS Power BI andExcel to generate actionable statistical reports and advanced visualizations. This work has uncoveredover 10 critical insights that continuously optimize marketing strategies and A/B testing, leading to a20% improvement in decision-making. I orchestrate data analysis and preprocessing of complexdatasets using Python and SQL, which has streamlined workflows and enhanced operationalefficiency by 30%. Collaborating with senior data science teams, I refine deep learning, unsupervisedlearning, and recommendation models, boosting model accuracy by 10% and ensuring an 85% projectsuccess rate. Additionally, I execute sentiment analysis on customer feedback using NLP techniques,identifying trends that increase customer retention by 10% and reduce complaints by 15%,significantly enhancing customer satisfaction.Achievements:- Identified and implemented actionable insights from data analysis, resulting in increased overallmarketing effectiveness.- Streamlined data processing workflows, leading to a reduction in time-to-insight for key businessmetrics.- Refined machine learning models and techniques, achieving improvements in model accuracy andbusiness outcomes. Show less