Hello, I'm

Amar Kumar

Engineer by training, data scientist by passion. I build intelligent systems that turn raw data into actionable insights — bridging the gap between complex analytics and real-world impact.

The Story So Far

Portrait of Amar Kumar
0+ Projects
0+ Certifications
0+ Years Learning

I've always been the person who asks "why?" one too many times. That curiosity led me from an engineering classroom to the world of data science, where every dataset tells a story waiting to be uncovered.

My engineering background gave me the discipline of structured problem-solving. But it was my first encounter with a messy, real-world dataset that truly lit the spark — the thrill of cleaning noise into clarity, of finding the signal that drives a decision. Since then, I've been building models, wrangling data, and chasing that feeling every single day.

I believe great data science isn't just about algorithms — it's about asking the right questions and communicating answers that people can act on. Whether it's predicting customer behavior or visualizing trends that shift strategy, I'm driven to make data useful.

Curious by Nature Detail-Oriented Always Learning Team Player Problem Solver

My Technical Toolkit

Programming

Python90%
SQL85%
R70%

Data & ML

Pandas / NumPy88%
Scikit-learn82%
TensorFlow65%

Visualization

Tableau80%
Power BI75%
Matplotlib / Seaborn85%

Tools

Git / GitHub85%
Jupyter Notebook92%
VS Code / Excel88%

Things I've Built

Customer Churn Prediction project screenshot

Built a classification model achieving 93% accuracy to identify at-risk telecom customers using ensemble methods and feature engineering on a 7,000-row dataset.

Customer Churn Prediction

ML model to identify customers likely to leave, enabling proactive retention strategies.

PythonScikit-learnPandasXGBoost
Sales Forecasting Dashboard screenshot

Interactive Tableau dashboard backed by a time-series ARIMA model, forecasting quarterly revenue for a retail chain with 94% directional accuracy.

Sales Forecasting Dashboard

Time-series analysis and interactive dashboard to project future sales trends.

PythonTableauARIMAStatsmodels
NLP Sentiment Analyzer screenshot

Fine-tuned a Transformer-based model on 50k+ product reviews, deploying a Flask API that classifies sentiment in real time with 91% F1-score.

NLP Sentiment Analyzer

Deep learning pipeline for real-time sentiment classification of product reviews.

PythonTensorFlowNLPFlask
Exploratory Data Analysis project screenshot

Comprehensive EDA on a public health dataset, uncovering key demographic correlations with interactive Plotly visualizations and statistical testing.

EDA on Public Health Dataset

In-depth exploratory analysis revealing hidden patterns in public health data.

PythonPlotlySeabornJupyter

My Journey

2024 — Present

Data Science Intern

TechAnalytica Solutions

Worked on predictive analytics pipelines, built ML models for client engagement scoring, and automated ETL workflows reducing manual processing time by 40%.

2023

ML Research Project Lead

University Research Lab

Led a 4-member team on a sentiment analysis research project using Transformer architectures. Published findings in the university journal and presented at a regional data science symposium.

2021 — 2025

B.Tech in Computer Science & Engineering

National Institute of Technology

Relevant coursework: Data Structures & Algorithms, Statistical Methods, Machine Learning, Database Management Systems, Linear Algebra, Probability & Statistics.

2021

Higher Secondary (XII) — Science

Delhi Public School

Scored 94% with a focus on Mathematics and Computer Science. Won the inter-school coding competition and led the school's tech club.

Credentials That Back It Up

Google Data Analytics

Professional Certificate

IBM Data Science

Professional Certificate

Kaggle Competitions

Top 15% — Titanic & Housing Prices

HackerRank Python

Gold Badge (5 Stars)

Deep Learning Specialization

Coursera — Andrew Ng

SQL Advanced

HackerRank Certification

Thoughts & Learnings

Where I share what I'm learning, building, and thinking about in the world of data.

Blog post illustration: building an ML model Machine Learning

How I Built My First ML Model from Scratch

A walk through the messy, frustrating, and ultimately rewarding process of going from a CSV file to a working prediction model — no magic, just persistence.

Read More
Blog post illustration: Python libraries Python Tips

5 Python Libraries Every Data Scientist Should Know

Beyond the obvious Pandas and NumPy — here are five lesser-known libraries that have saved me hours of work and made my analyses cleaner.

Read More
Blog post illustration: confusion matrix explained Career

Making Sense of Confusion Matrices

Precision, recall, F1, specificity — it all made zero sense until I reframed it with a real example. Here's the explanation I wish I had when I started.

Read More

Let's Build Something with Data

Have an idea, a question, or just want to say hello? I'd love to hear from you.

Location

India

Find me on