Hamza Uzzaman

Kolkata, IN · B.Tech · Computer Science & Engineering
ml cv backend agents fastapi deep learning

who I am

Third-year Computer Science student focused on building deployable machine learning systems, computer vision applications, and intelligent backend architectures.

I have a builder mindset — I prefer shipping projects over reading theory. Most of my learning comes from breaking my own models and figuring out why they failed. That's where real engineering happens.

Currently pushing toward advanced ML, deep learning architectures, and real-world deployment pipelines.

projects

End-to-end regression pipeline — data cleaning, outlier removal, feature engineering (location encoding, area scaling). Trained and compared Linear Regression, Decision Tree, and Random Forest with RMSE and cross-validation evaluation. Simulates a full real-world ML workflow from data to inference.

PythonPandasNumPyScikit-learn

Real-time hand landmark detection and finger tracking with high accuracy. Built for extensibility — usable for gesture control, virtual mouse, and HCI systems. Optimised frame processing for smooth real-time performance.

PythonOpenCVMediaPipe
⚖️ Obesity Checker
cv health wip

Vision-based system to estimate body metrics and classify obesity levels. Uses image processing to extract body proportions and features, with planned ML integration for BMI estimation. Focus on production-ready health-assist deployment.

OpenCVPythonML
🧠 ML Algorithm Sandbox
ml fundamentals

From-scratch implementations of core ML algorithms — Logistic Regression, KNN, Linear/Ridge/Lasso Regression — alongside full data preprocessing pipelines and model evaluation. Focused on deeply understanding model behaviour, not just using libraries.

PythonScikit-learnNumPy

tech stack

Languages

Python
Java
C
C#
Kotlin
JavaScript
HTML / CSS
PHP

ML / AI

Scikit-learn
PyTorch
TensorFlow
Keras
NumPy
Pandas
SciPy
OpenCV
Plotly
MLflow

Backend & Frameworks

FastAPI
Django / DRF
Flask
Spring
React
Streamlit
GraphQL
Express.js

Cloud & Infra

AWS
Azure
Heroku
Vercel
Firebase
Nginx
Git / GitHub
Cloudflare

Databases

MongoDB
MySQL
PostgreSQL

Most of my learning comes from breaking my own models and figuring out why they failed — that's where the real engineering happens.