Melodia Academy
Web AppMusical institution management platform (Laravel + SQLite3).
👋 Hi, I’m Jehan — building software & exploring applied AI.
A clean, modern portfolio
with a research segment.
CSE student at United International University (2022–2026). I build full-stack apps using Django/React and experiment with machine learning using scikit-learn / PyTorch.
CGPA
3.93 / 4.00
Focus
Full-Stack + ML
Tools
Git / GitHub
About
I enjoy building real products and documenting experiments. I care about clean UI, readable code, and measurable results.
What I do
- Full-stack apps (Django / React / PHP / Next.js)
- Databases (SQLite3, MySQL)
- ML experiments (scikit-learn, TensorFlow/Keras, PyTorch)
- Version control (Git, GitHub)
Highlights
Scholarships
100% & 50% academic scholarship.
Math Talent Probe
Rank 390 (Global Mathematical Talent Probe 2017/18).
Projects shipped
Tour, nutrition, and real-estate collaboration platforms.
Tech Stack
A quick overview of tools I commonly use.
C / C++
Java
Python
HTML5
CSS3
JavaScript
TypeScript
React
Tailwind
Django
Laravel
Next.js
PHP
SQLite3
MySQL
Scikit-learn
TensorFlow
Keras
PyTorch
Git
GitHub
Projects
Click a project to view details in a modal.
NutriPick
ML + WebNutrition planner with ML features.
BuildIT
PlatformReal-estate collaboration & workflow.
Explore Plus
Web AppTour management platform (PHP + MySQL).
Research Papers
A research segment with filters. These are dummy/sample papers you can replace with your real work later.
NutriPick: Lightweight Meal Recommendation with Feature-Aware Modeling
Preprint • 2026Keywords: recommendation, nutrition, classification • Dataset: synthetic + user logs (simulated)
We explore a lightweight recommendation approach for daily meal planning using feature-aware modeling. The system combines nutrition targets and constraints, producing practical meal suggestions. We report improvements over a baseline heuristic using offline evaluation metrics (precision@k, coverage).
ExplorePlus: Demand-Aware Tour Package Ranking using Simple Forecasting
Workshop Submission • 2025Keywords: tourism, forecasting, clustering • Focus: ranking tour packages by predicted demand
This paper proposes a simple demand-aware ranking method for tour package recommendation. We combine seasonal signals with user preference categories and evaluate ranking consistency over time. The method is designed to be easy to implement in typical LAMP-style web applications.
BuildIT: A Collaboration-First Workflow for Small Builder Teams
Technical Report • 2024Keywords: software engineering, collaboration, workflow • Focus: requirements + UX + system design
We present a collaboration-first workflow model for small builder teams and outline a platform design supporting task tracking, file sharing, and job discovery. The report includes UX decisions, database design, and a modular backend architecture that scales with small teams.
Contact
Let’s connect for internships, collaboration, or project work.