Art
Turning real-world data into insights, models, and scalable apps
Click to change geometry
I'm Shivam Chavan, a tech enthusiast with a background that bridges both engineering and computer science. I completed my Bachelor's in Mechanical Engineering, where I developed a strong foundation in problem-solving and analytical thinking. Driven by a growing passion for technology, I pursued a Master's in Computer Science at George Washington University, where I’ve been sharpening my skills in data science, full-stack development, and cloud technologies. Over time, I’ve developed proficiency in tools like Python, React, SQL, MongoDB, GCP, and AWS, picking up new skills through hands-on learning, experimentation, and continuous curiosity. Outside the world of code, I’m deeply passionate about art. You’ll find a dedicated section for my artwork at the end of this portfolio. I’m currently open to full-time opportunities in roles such as Software Engineer, Data Analyst, Data Scientist, Backend Developer, or Full-Stack Developer — feel free to connect with me!
HTML
CSS
Pandas
KMeans
JavaScript
Python
React
Node.js
MongoDB
SVM
SQL
Power BI
VBA
GCP
AWS
Docker
Git
GitHub
Oracle APEX
Scikit-learn
Machine Learning
NumPy
Data Visualization
Flask
APIs
BigQuery
ETL Pipelines
Gradient Boost
Databricks
Jupyter
AdaBoost
Looker Studio
Leaflet
XGBoost
R Shiny
Naive Bayes
Built a scalable ETL pipeline using Mage.ai on Google Cloud Platform to process and analyze NYC Yellow and Green Taxi trip data. Leveraged BigQuery for data warehousing, GCP Storage for ingestion, and Looker Studio for interactive dashboards. The pipeline automated transformations, enabling trend analysis and insights across millions of records.
Developed a full-stack inventory optimization tool using ReactJS, Node.js, and MongoDB, designed for small restaurants to forecast demand and reduce food waste. Integrated Databricks to apply Stochastic Gradient Descent for 7-day demand prediction, helping users make data-driven purchasing decisions.
Created an interactive data visualization tool using R Shiny and Leaflet, showing rental price trends across U.S. states and counties. The app allows users to explore geospatial rent data with filters by region and time period, making housing affordability insights easily accessible to users.
Designed a complete machine learning pipeline using Python, Pandas, Scikit-learn, and SMOTE to detect fraudulent claims. Trained multiple classification models, evaluated with metrics like precision, recall, and ROC-AUC, and optimized the best-performing model for real-world applicability.
Developed a responsive web app using React.js, Redux Toolkit, and ReactStrap for a modern UI/UX. The app supports user registration, meal browsing, cart management, and order placement, simulating a real-world meal delivery platform.
Developed a flight delay prediction model using Decision Tree, AdaBoost, Gradient Boosting, and XGBoost, with performance evaluation across key metrics. Built a Naive Bayes model for chronic heart failure prediction with strong accuracy, and explored emotion recognition using KMeans for feature extraction and SVM for classification.
See more on GitHub
The George Washington University
Washington, DC
Black Rocket Productions
Fairfax, Virginia
PSCWP
Washington, DC
OSS Air Management Pvt. Ltd.
Mumbai, India
Pillai College of Engineering
Panvel, India
MCM
Navi Mumbai, India
See more on LinkedIn
Feel free to get in touch with me!
shivamchavan05@gmail.com