Antarik Biswas

Data Scientist

Resume

Hello, I am

Antarik Biswas.

2

Machine Learning
Projects

About Me

Hi, I’m Antarik Biswas, a Computer Science graduate with a strong passion for working with data and solving real-world problems through technology.

🛠️ Technologies & Tools
🔷Programming: Python, SQL (MySQL)
🔷 Libraries: Pandas, NumPy, Matplotlib, Seaborn
🔷 Machine Learning: Scikit-learn
🔷 Web & APIs: Streamlit, FastAPI
🔷 Version Control: Git & GitHub

💼 What I Bring
🔵 Hands-on experience in data extraction, transformation, and visualization
🔵 Ability to translate business needs into data-driven solutions
🔵 Strong analytical mindset with a focus on accuracy and efficiency
🔵 Commitment to continuous learning and knowledge sharing

Key Skills

PowerBI

Python

MySQL

Mlflow

Github

Exploratory Data Analysis (EDA)

Data Visualization

Statistical Data Analysis

Machine Learning

My Projects

Health Insurance Premium Predictor

Domain/Function: Healthcare

Credit Risk Evaluator

Domain/Function: NBFC Credit Risk Modelling

My Experience

Infosys Bangalore

System Engineer (Full-Time) Jun 2022 - Feb 2024

Improved reporting speed and effectiveness by 30% by integrating data from multiple sources and enhancing data quality and reliability. Validated and cleaned large datasets (100K+ records) using Excel and SQL to ensure data accuracy and consistency.
Resolved front-end issues in React.js applications, including layout mismanagement and load speed optimization, ensuring smooth functionality and enhanced user experience with a 95% issue resolution rate.

Awards & Certificate

Get Job Ready: Power BI Data Analytics for All Levels 3.0

SQL Beginner to Advanced For Data Professionals

SQL for Data Science

Math and Statistics For AI, Data Science

Master Machine Learning for Data Science & AI: Beginner to Advanced

Python: Beginner to Advanced For Data Professionals

Google Data Analytics

Let's Connect

Feel free to get in touch with me. I am always open to discussing new projects, creative ideas or opportunities to be part of your visions.

https://www.linkedin.com/in/antarik1109/

GitHub

https://github.com/Antarik-80

Machine Learning Project

Health Insurance Premium Predictor

Domain/Function: Healthcare

Developed a Streamlit-based web app to predict health insurance premiums using machine learning. Included end-to-end preprocessing, feature engineering, and real-time prediction capabilities.

View Project
on GitHUB

View Project
on

Prediction for Young user

Prediction for Old user

Project Details

Manually estimating health insurance premiums can be both inaccurate and time-consuming. To address this, I developed a Health Insurance Premium Prediction system powered by machine learning, deployed as a fully responsive Streamlit web app.

The solution incorporates a complete ML pipeline using Pandas, Scikit-learn, and tailored feature engineering techniques, enabling real-time predictions based on users' demographic and health information.

🔑 Key Highlights:
⦿ Implemented 10+ pre-processing steps, including VIF-based feature reduction and outlier removal
⦿ Reduced prediction error for younger users from 73% to 2% through age-based model segmentation
⦿ Achieved an R² score of 0.988 on the test dataset
⦿ Delivers real-time predictions in under one second

This project demonstrates end-to-end machine learning in the insurance technology space—blending robust data preparation, intelligent model optimization, and seamless frontend deployment.

Machine Learning Project

Credit Risk Evaluator

Domain/Function: NBFC Credit Risk Modelling

Developed a credit risk prediction system using machine learning to assess loan default probability, offering real-time credit scoring and risk categorization through a responsive Streamlit web app.

View Project
on GitHUB

View Project
on

Poor Rating

Excellent Rating

Project Details

Manually assessing loan risk can be both time-consuming and inconsistent. To solve this, I developed a machine learning–powered Credit Risk Evaluator that predicts the likelihood of loan default using borrower profiles and financial history.

The solution is built with Streamlit for an interactive frontend and integrates a complete preprocessing and modeling pipeline using Scikit-learn. Key engineered features such as the Loan-to-Income (LTI) Ratio and Average Days Past Due (DPD) enhance model performance, especially recall on high-risk applicants. The system also supports real-time credit scoring for new applicants.

🔑 Key Highlights:
- Predicts loan default risk using borrower and financial data
- Converts model output to a risk score (300–900) for easy interpretation
- SMOTE Tomek resampling and Optuna-based hyperparameter tuning for enhanced recall
- Fully functional Streamlit web app for real-time scoring and decision support