Rounak Kumar
Rounak Kumar

About Me

I'm an Integrated MSc student in Quantitative Economics & Data Science at BIT Mesra with a strong focus on building production-grade machine learning systems.

My work spans statistical modeling, machine learning, and MLOps, with hands-on experience deploying scalable ML pipelines using Docker, GitHub Actions, and Azure.

5+

end-to-end ML Projects

100+

DSA Questions

Joint President

Data Science Society

Coding Statistics

My journey in competitive programming and open source

GitHub Streak

GitHub Streak
View Profile

Most Used Languages

Python 65%
Jupyter Notebook 20%
Dockerfile 5%
CSS 3%
View Repositories

What I Bring to the Table

Core competencies that drive my work

Data Science

Data-Driven Modeling & Production ML

ML Engineering

End-to-End ML Systems

MLOps Expertise

Docker & CI/CD Pipelines

Cloud Deployment

Azure & AWS Platforms

Skills & Technologies

Technologies I work with

Languages

Python
R R
SQL
C++

ML / Data Science

Scikit-learn
TensorFlow
PyTorch
LightGBM LightGBM
CatBoost CatBoost
XGBoost

Backend & Databases

Flask
FastAPI
Streamlit
MongoDB
MySQL

MLOps & Cloud

Docker
GitHub Actions
Azure
AWS
MLflow MLflow
Comet ML Comet ML

Data Analytics

Power BI Power BI
Snowflake Snowflake
Seaborn Seaborn
Pandas Pandas
SQL

My Projects

Here are some of my recent works

🎌 Hybrid Anime Recommendation System

Built a scalable hybrid recommendation system processing 70M+ user–anime ratings, combining collaborative and content-based filtering for improved recommendation accuracy. Features automated large-scale data ingestion using Azure Blob Storage and embedding-based hybrid recommender using Keras.

Python Keras Comet ML Docker GitHub Actions Azure Blob Storage Azure Web App
Render Live Demo

🛡️ Network Security System

Designed and deployed a production-grade anomaly and phishing detection system with secure, automated workflows and real-time prediction capabilities. Features MongoDB Atlas–based ingestion with schema validation and MLflow integration for experiment tracking.

Python MongoDB Atlas MLflow FastAPI Docker GitHub Actions Azure Containers
Render Live Demo

🧠 Student Performance Predictor

Deployed an end-to-end pipeline that forecasts student scores (R² ≈ 0.87) with automated ingestion, data quality checks, feature scaling/encoding, and model retraining. CI/CD (GitHub Actions) ships Dockerized builds to Azure Web App, with drift-ready hooks for future monitoring.

Python Scikit-learn CatBoost Pandas Docker GitHub Actions Azure Web App
Live Demo

Get In Touch

Feel free to reach out for collaborations or just a friendly chat

"Code that talks to data, systems that talk to users."