About Me

👋 Hey, I'm Hrushikesh!

I'm a Data Science | Machine Learning | Analytics enthusiast, Currently pursuing my Master’s in Data Science at George Washington University. I hold a Bachelor's degree in Computer Science and Engineering with a specialization in Artificial Intelligence from Vellore Institute of Technology (VIT).

With a strong foundation in computer science and a passion for transforming data into impactful stories, I enjoy solving real-world problems through machine learning, analytics, and visualization. Beyond models and dashboards, I’m a huge cricket enthusiast 🏏 and love experimenting in the kitchen 🍳.

Education

Graduate Education

Master's in Data Science

Institution: George Washington University

Degree: Master's in Data Science

GPA: 4.0/4.0

Courses:Data Warehousing, Data Mining, Machine Learning, Data Visualization,Cloud Computing,Algorithms For Datascience.

Undergraduate Education

B.Tech in CSE (AI)

Institution: Vellore Institute of Technology

Degree: B.Tech in Computer Science & Engineering (AI Specialization)

GPA: 3.63/4.0

Courses: Data Structures and Algorithms, Artificial Intelligence, Deep learning, Linear Algebra, Advanced Data Analytics, Database Management Systems.....

Projects

🌡️📈 Forecasting Temperatures using Time Series and Advanced ML Techniques.

George Washington University, US | Spring 2025

Developed a robust forecasting system using ARIMA, SARIMA, Random Forest, XGBoost, and LSTM, leveraging both statistical and deep learning approaches for comparison.

Engineered advanced features like lag values, rolling averages, seasonal encodings, and week/day-based time markers to capture complex temporal patterns.

Performed seasonal decomposition (STL) to isolate trend, seasonality, and noise, improving model interpretability and preprocessing strategy.

Conducted city-wise forecasting for 400+ cities using optimized pipelines and visualized results via interactive dashboards (e.g., Plotly/Streamlit/Tableau).

LSTM architecture used 2 hidden layers with dropout and standardized inputs to handle sequence dependencies; achieved better R² despite overfitting challenges.

Evaluated models using RMSE, MAE, and R², demonstrating XGBoost’s superior balance of accuracy and efficiency.

⚡📊 Energy Consumption Forecasting using AWS

George Washington University, US | Fall 2024

Built ML models (Random Forest, AutoML) on AWS to predict energy usage from 1973 to 2024.

Achieved R² of 0.95 and RMSE of 0.209 using AutoML, identifying temperature and month as key drivers.

Visualized energy demand trends using AWS QuickSight to support non-technical decision-making.

📊🌍 Global Energy Trends and Impact

George Washington University, US | Spring 2024

Connected four datasets (global electricity use, regional disparities, energy source mix, CO₂ emissions) into a single data-driven narrative using DataWrapper and custom HTML.

Crafted a compelling visual story showing how global energy consumption is rising, still reliant on fossil fuels, and directly linked to increasing CO₂ emissions.

Highlighted global inequalities in electricity access and clean energy use using clear and responsive interactive charts.

Focused on visual clarity and argument-building, not just aesthetics or tools, aligning with real-world data storytelling principles.

📊🚗 Electric Vehicle (EV) Adoption Analysis Across U.S. Regions

George Washington University, US | Fall 2023

Explored regional EV adoption patterns across the U.S. using a large real-world dataset, identifying growth trends and regional disparities using SMART questions.

Cleaned and preprocessed 12,000+ records, mapped states to U.S. regions, and engineered features like EV type and year to prepare the data for meaningful insights.

Visualized adoption trends with Seaborn and Matplotlib, revealing hotspots like California and highlighting policy-driven uptake in the Northeast and West.

Validated hypotheses using ANOVA, Chi-square, and T-tests to uncover statistically significant differences between regions and EV types.

Built predictive models (OLS & Logistic Regression) to forecast high vs. low EV adoption zones, achieving 81% accuracy using demographic and vehicle factors.

☕📈 Starbucks Offer Success Prediction

George Washington University, US | Fall 2023

Grouped 17,000+ Starbucks users using K-Means clustering to better understand customer types and behaviors.

Predicted success of different promotional offers using Logistic Regression and Decision Trees (AUC up to 0.97).

Designed a custom "Offer Success" metric to help Starbucks tailor their BOGO, Discount, and Informational campaigns.

🧠👁️ Facial Feature Detection Using CNN

Vellore Institute of Technology, India | 2022

Created a real-time face analysis tool that predicts age, gender, emotion, and counts eye blinks using CNN.

Used MediaPipe (468 facial landmarks) + OpenCV for precise face tracking and feature extraction.

Achieved over 90% accuracy for age and emotion prediction, useful in HCI and biometric authentication.

Experience

Data Scientist Intern – ZettaMine Labs Pvt. Ltd.

May 2024 – Aug 2024

Developed a patient risk stratification model using Random Forest (F1-score: 88%) to help clinical teams prioritize high-risk cases for early intervention.
Engineered 15+ clinical features from vitals, diagnosis codes, and treatment history to enhance prediction accuracy and reliability.
Built modular data pipelines using Python, SQL, and Git that Reduced preprocessing time by 30%.
Delivered real-time insights to hospital decision-makers by integrating model outputs into Power BI dashboards and SAP Analytics Cloud reports are used across 3+ departments.

Data Scientist Intern –Bizom

Sep 2022 – Jul 2023

Built a machine learning model in Python that could predict which users were likely to stop using the app (user churn), with 85% accuracy.
Used data from notifications sent to users and turned it into Simple charts and graphs for the team to easily understand.
Shared those visual insights every 2 weeks with non-technical teams (like marketing), helping them make better decisions.
Created a model to find the best time to send notifications to users, which increased the number of people clicking on them by 15%.
Helped reduce how often users felt annoyed by notifications (called notification fatigue) by 10%.
Took initiative in improving the models even after deployment, ensuring they remained accurate and useful.

Skills

Languages

Frameworks / Tools

Developer Tools

Cloud / Big Data

Python Java SQL HTML/CSS R

scikit-learn Keras PyTorch TensorFlow OpenCV Tableau Power BI MongoDB Neo4j MATLAB Seaborn ggplot2 Matplotlib Pandas NumPy Statsmodel MediaPipe Plotly D3.js YOLO AutoML TabPy

GitHub Git Jupyter Notebook Google Colab Eclipse NetBeans RStudio Visual Studio

AWS S3 AWS Lambda AWS Glue Amazon SageMaker Amazon QuickSight Microsoft Azure Kubernetes Google Cloud Platform

Certifications

Microsoft Certified: Azure Data Scientist Associate

Microsoft

Issued: August 2025

Expiry: August 2026

Credential ID: 2D9F089EE725ECD0

Click to view certificate

AWS Certified Data Engineer - Associate

AWS

Issued: July 2025

Expiry: July 2028

ID: f32d75fe5b984e5694d74e1c3efa6cf4

Click to view certificate

Oracle Cloud Infrastructure 2025 Certified Data Science Professional

Oracle

Issued: July 2025

Expiry: July 2027

Click to view certificate

Oracle Cloud Infrastructure 2025 Certified Generative AI Professional

Oracle

Issued: July 2025

Expiry: July 2027

Click to view certificate

Certified AI Specialist

Salesforce

March 2 2025

ID: 5839250

Click to view certificate

Red Hat OpenShift Development I: Introduction to Containers with Podman

Redhat

Dec 9 2024

Click to view certificate

Red Hat OpenShift I: Containers & Kubernetes

Redhat

Dec 10 2024

Click to view certificate

Introduction to R

Datacamp

Sep 9 2023

Click to view certificate

Developing Soft Skills and Personalitys

NPTEL

Sep 25 2022

Click to view certificate

Machine Learning Masters

iNeuron

Feb 2021

Resume

Download PDF

Contact Me

I'll be glad to answer your questions!