Tejaswini Parlapalli

Tejaswini Parlapalli

Welcome

Hello,

I'm a data enthusiast with 6 years of experience in analytics, focused on building and deploying machine learning models. Feel free to explore my Projects which showcase my data science and analysis work. I enjoy applying both analytical and creative thinking to transform complex data into clear, actionable insights and when I step away from the keyboard, you can usually find me with a good book, exploring the local food or singing.
Thanks for dropping by!

Download Resume

Experiences

Rivian Logo
Apr 2023 - Present

Rivian

Sr. Data Scientist

Implemented predictive machine learning models (Linear Regression, Random Forest, Gradient Boosting) to forecast resource shortages, reducing operational downtime by 5% and improving overall process efficiency by 12%. Optimized demand forecasting with predictive analytics (SVM, Bayesian methods, Linear Regression), achieving a 15% reduction in excess inventory. Developed Alteryx workflows delivering key KPIs to stakeholders, reducing production delays, bottlenecks and enabling faster, data-driven decisions.

Data Scientist

Enhanced allocation of shared line-side inventory using advanced ML classification models (XGBoost, LightGBM), boosting utilization by 20% and ensuring stock distribution matching to teams’ requirements. Implemented a Reinforcement Learning (Q-Learning) system to autonomously optimize inventory and production planning – a project recognized with a Leadership Innovation Award. Streamlined data from multiple systems (SAP & MES) into cloud platforms (AWS Redshift, Snowflake, Databricks) and created executive-level Tableau dashboards, improving visibility and decision-making.

IU Marketing Analytics Logo
Aug 2022 - May 2023

IU Marketing Analytics

Data Scientist

Implemented advanced Alteryx ETL model automations to enhance API connector performance, reducing data processing time and enabling seamless integration into Big Query. This initiative saved 10 man-hours per week and streamlined workflow efficiency. Additionally, optimized pipeline development strategies within marketing funnels by leveraging Big Data tools like Hive and Spark, leading to a significant 15% reduction in data maintenance costs. Furthermore, implemented GA4 processes to track user behavior and improve marketing effectiveness. By translating these insights into Looker Studio visuals, decision-making was facilitated, resulting in a notable 2% increase in website conversions.

Indiana University Logo
Aug 2021 - Apr 2022

Indiana University

Research Assistant

Pioneering a data-driven approach, spearheaded the integration of state-of-the-art Ml & DL techniques into our research framework. This involved the development and deployment of ML and DL classification models, alongside neural network architectures, to accurately determine the age of real-world images with an precision rate of 78.3%. Additionally, leveraging NLP algorithms such as FastText, developed a failure mode prediction systems, resulting in a substantial 53% enhancement in quality metrics. Moreover, by employing NLP techniques for feature engineering, extracted valuable insights from textual data linked to images.This modified our models with crucial features, significantly improving their predictive capabilities.

EMC Insurance Logo
Jun 2022 - Aug 2022

EMC Insurance Group

Data Scientist

By implementing regression models such as Random Forest, Gradient Boost, and XGBoost, alongside meticulous feature engineering and optimization of hyperparameters on AWS and Snowflake data, I achieved an impressive 60.1% damage prediction rate. Delving into extensive research and deep analysis of large datasets, I identified crucial trends and integrated BERT, leading to a notable 10% improvement in the accuracy of insurance analysis. Moreover, through the development of interactive dashboards and visualizations using Power BI, stakeholders gained enhanced insights, aiding in the identification and rectification of faulty investments totaling $260k.

Deloitte Logo
Dec 2019 - Jul 2021

Deloitte

Business Intelligence Analyst

Played a pivotal role in developing an Oracle Cloud RDBMS framework and implementing data analysis procedures, resulting in a seamless migration .Also by leveraging MySQL, Python, and Bash scripts, optimized data processes, ultimately reducing migration time by 3%.Utilizing statistical methods conducted data trend analysis with an impressive accuracy rate of over 85%, providing valuable insights for strategic decision-making.

Amazon Logo
Dec 2019 - Dec 2019

Amazon

Data Scientist

Played a pivotal role in developing an Oracle Cloud RDBMS framework and implementing data analysis procedures, resulting in a seamless migration .Also by leveraging MySQL, Python, and Bash scripts, optimized data processes, ultimately reducing migration time by 3%.Utilizing statistical methods conducted data trend analysis with an impressive accuracy rate of over 85%, providing valuable insights for strategic decision-making.



Things I've Built

Age Gender detection system using CNN and RNN deep learning techniques

The Age, Gender, and Ethnicity Detection System is a deep learning framework designed to classify age, gender, and ethnicity from facial images. Utilizing CNN and RNN models, this system aims to provide accurate and reliable results for applications in airports, police stations, and personalized recommendation systems.

Analysis of Airline Passenger satisfaction

The goal is to identify key factors contributing to passenger satisfaction. Using R Studio, machine learning algorithms, and data visualizations, the analysis highlights the impact of inflight entertainment, ease of online booking, and travel type. Linear Discriminant Analysis (LDA) and Quadratic Discriminant Analysis (QDA) were the most accurate models.

Zonal Accident Risk Prediction system

The Zonal Accident Risk Prediction project analyzed road accident data to identify patterns and contributing factors. Using linear regression and random forest models, the project predicted factors like speed limits and fatalities to recommend safety measures. Limitations included missing predictors for road surface and weather conditions.

The IUCommuterPass (IUCP) web app connects IUPUI students for carpooling, reducing transportation costs and environmental impact. Features include trip cost calculation, check-in, and messaging. The project promotes affordable transportation, reduces pollution and traffic congestion, and fosters social connections among students.

John snows cholera map visualization using D3.js

Dr. John Snow's map visualized the 1854 cholera epidemic in London, revealing patterns and mitigating effects. Interactive elements included a timeline chart, map, and pie charts showing death variations by age and gender. Key findings highlighted higher death rates among the elderly and children, with contaminated water pumps as hotspots.

End-to-End MLOps Fraud Transaction Scoring API

An end-to-end MLOps project for real-time credit card fraud detection, project involves architecter and deployment of a scalable API leveraging a trained LightGBM model with an automated workflow via a CI/CD pipeline using GitHub Actions and Docker, ensuring continuous deployment on cloud app render.

AI Code Review Agent

An AI-powered Code Review Assistant using Streamlit and OpenAI that automatically analyzes code for best practices, potential bugs, and readability improvements — helping developers accelerate reviews and write cleaner code


Education

College

Indiana University Purdue University Indianapolis

Masters, Applied Data Science

GPA 3.8

Courses: Database Management, Programming - Data Science, Cloud computing, Web Database development, Data Analytics with R, Machine Learning, Deep Learning & AI, Statistical Learning and Data visualization.

College

Jawaharlal Nehru Technological University

Bachelors, Electronics Engineering

GPA 4.0

Courses: C, Java ,Matlab, Python, Database management,Data processing at scale, Advanced computer networks,Operating systems

Skills



Gmail GitHub LinkedIn Leetcode