Kenean Dita Meleta
Software Engineer | Container Guru | ML Enthusiast
Powerful Narrative
"Everything should be made as simple as possible, but no simpler."
— Albert Einstein
Untangling a complex problem reveals its core truth. A powerful strategy is not a dense document, but a clear, actionable path forward. It distills chaos into clarity, empowering focused action and achieving profound results with apparent ease.
About Me
Resume
Focused on achieving project goals through clear planning, effective teamwork, and attention to detail. Skilled in managing tasks, solving problems, and adapting to changing needs to deliver quality results on time.
Download ResumeMy Tech Stack
Python
Django
Flask
FastAPI
TensorFlow
PyTorch
Keras
Scikit-learn
Streamlit
Jupyter
MySQL
Git
GitHub
Docker
KubernetesProjects
SentriX
A minimal machine learning project that analyzes and predicts cryptocurrency market movements.
TensorflowTorchStreamlitScikit-learnMatplotlibDockerMLCrypto
Fancy-Agent
A command-line interface (CLI) tool designed to assist with code generation, syntax highlighting, and interactively prompting users to execute commands.
PythonRichGoogle-GenAIAIDeveloper Tools
Vendly
A comprehensive web platform built with Django and Tailwind CSS, this site offers a seamless, modern shopping experience across diverse categories. Vendly has clean, responsive design and optimized performance.
DjangoTailwind CSSSQLliteDockerEcommerceRetail
GradeCast
A machine learning web application built with Flask that predicts student performance based on input data.This project showcases practical skills in data preprocessing, model training, evaluation, and deploying ML models using Flask for real-time predictions.
FlaskScikit-learnJupyter NotebookDockerMLEducation
Gitlog-CLI
A lightweight command-line tool that connects to the GitHub API to retrieve and display a user’s most recent public activity. It shows actions such as commits, pull requests, issues, and repository events in a clean, readable format directly in the terminal.
UrllibJsonRichDeveloper ToolsTerminal Utilities
Fidel-Vision
A Deep Learning + Streamlit web app for recognizing handwritten Amharic fidel characters (34 root groups × 7 orders). It uses a CNN model built with TensorFlow/Keras, trained on a custom dataset of handwritten Amharic letters, and serves predictions through a Streamlit interface.
StreamlitKerasTensorFlowPandasDockerDeep LearningAmharic Language