Moustafa Mohamed

Moustafa Mohamed

AI Developer | Specializing in Machine Learning, Deep Learning & LLM Engineering

30+

Projects

10+

Certifications

1

Published Library

View Projects Download Resume

About Me

I'm a second-year Software Engineering student specializing in Artificial Intelligence and Data Science. My expertise spans machine learning, deep learning, and large language models, with a passion for building intelligent systems that solve real-world problems.

Background

My technical journey began with C/C++, establishing strong foundations in algorithms and system-level programming. I've since mastered Python for AI/ML development, along with frameworks like TensorFlow, PyTorch, and Keras. My recent focus has been on LLM engineering, generative AI, and developing production-ready data science solutions.

Beyond coding, I'm passionate about the entire data pipeline - from ETL processes to deploying machine learning models. I've developed expertise in building end-to-end AI systems, including my published Python library for streamlined data cleaning and EDA.

Education

  • Bachelor's in Software Engineering
    Istanbul Topkapi University (2023-Present)
    Specialization: Artificial Intelligence & Data Science

Key Achievements

  • Published PyPI package (datacmp) for data cleaning
  • Developed multiple production-grade AI models
  • Certified in Deep Learning and LLM Engineering
  • Created LLM Tools Suite with live demo

Experience

AI Intern

BLUESENSE · Internship
Jun 2025 - Jul 2025 · 2 mos
Vancouver, British Columbia, Canada · Remote

Contributed to the research, development, and optimization of AI-powered solutions, with a focus on computer vision and deep learning emphasizing model performance, interpretability, and reproducibility.

Key Responsibilities:

  • Designed and fine-tuned convolutional neural network (CNN) architectures for image-based analysis tasks
  • Applied advanced data preprocessing, augmentation, and model training techniques to improve performance and generalization
  • Leveraged transfer learning methodologies to accelerate model development and enhance accuracy

Core Technical Skills:

Transfer Learning TensorFlow Keras Convolutional Neural Networks (CNN) PyTorch OpenCV Git YOLOv8

Kaggle Expert & Contributor

Kaggle · Self-employed
Feb 2025 - Present · 6 mos
Remote

Recognized as a Notebooks Expert on Kaggle, ranked in the top 5% globally out of 56,000+ contributors.

Key Achievements:

  • Earned 5 Bronze Medals for published notebooks
  • Awarded 1 Gold, 3 Silver, and 34 Bronze Medals for valuable contributions to discussions
  • Developed notebooks demonstrating real-world applications of AI and LLMs
  • Strengthened expertise in Python, applied machine learning, and deep learning frameworks

Core Activities:

  • Contributed actively to the Kaggle community across Discussions, Datasets, Notebooks, and Competitions
  • Shared reusable code and built ML notebooks
  • Demonstrated initiative and knowledge sharing within one of the world's largest data science communities

Technical Skills:

Python Machine Learning Deep Learning Data Analysis Large Language Models (LLM)

Technical Skills

AI & Machine Learning

Deep Learning

Neural Networks

Generative AI

NLP

LLM Engineering

Prompt Engineering

LLaMA

Pipelines

RAG

Fine-Tuning (LoRA/QLoRA)

AI Agents

Function Calling

Code Optimization

Multi-modal AI

Data Science

Data Analysis

Data Cleaning

Data Visualization

Machine Learning

EDA

ETL

Plotly

Power BI

Programming & Tools

Python

PyTorch

TensorFlow

Keras

C/C++

JavaScript

SQL

Git

GitHub

PyPI

API Dev

OOP

HuggingFace

LangChain

Gradio

Model Deployment

Published Library

datacmp

PyPI

A Python library for streamlined data cleaning and exploratory analysis. Features automatic column standardization, missing value handling, and dataset summarization with YAML-based configuration.

Data profiling & visualization
Automated column cleaning
Missing value handling
YAML configuration
Command line interface
Export capabilities
PyPI Page GitHub
datacmp library screenshot

Projects

LLM & Generative AI Projects

LLM Tools Suite

Integrated collection of AI-powered tools including blog generation, SQL query builder, document summarization, and code explanation powered by Gemini and LLaMA 3.2.

Gemini API LLaMA 3.2 Streamlit LangChain FAISS
Live Demo
View Code

AI Chatbot

A professional AI chatbot interface powered by Google's Gemini API, featuring: - Real-time conversational AI - Session management - Responsive web interface - File upload capability - Markdown rendering

Gemini API Prompt Engineering LLMs
View Code

Machine Learning Projects

Credit Card Fraud Detector

XGBoost model achieving 99.96% accuracy in detecting fraudulent transactions. Includes EDA, feature importance analysis, and model comparison with 5 different algorithms.

XGBoost Scikit-learn Pandas Matplotlib Imbalanced Data
99.96% Accuracy
View Code

Laptop Price Prediction

Regression model predicting laptop prices based on specifications. Features extensive EDA, feature engineering (PPI calculation), and deployed Streamlit app for real-time predictions.

Linear Regression Random Forest XGBoost Streamlit Feature Engineering
View Code

Diabetes Prediction

Comparative analysis of multiple classifiers on medical data to predict diabetes. Includes comprehensive EDA with dark-themed visualizations and model performance evaluation.

Logistic Regression SVM Decision Tree Random Forest Seaborn
View Code

Deep Learning Projects

Fruit & Vegetable Image Recognition

CNN model classifying 36 different fruits/vegetables with 92% accuracy. Implemented image augmentation, transfer learning, and visualization of model performance.

TensorFlow Keras CNN Image Augmentation Computer Vision
92% Accuracy
View Code

Breast Cancer Prediction (Neural Network)

PyTorch-based feedforward neural network for tumor classification. Achieves high accuracy in predicting malignant vs benign cases with detailed performance metrics.

PyTorch Neural Network BCELoss Medical AI Scikit-learn
View Code

Fashion MNIST Classification

CNN architecture enhanced with Batch Normalization and Dropout for classifying fashion items. Achieves 91-93% accuracy with futuristic dark-themed visualizations.

TensorFlow Keras BatchNorm Dropout EarlyStopping
93% Accuracy
View Code

Certifications

LLM Engineering: Master AI & Large Language Models

Udemy

Issued May 2025 View Certificate

Deep Learning A-Z 2025

Udemy

Issued Feb 2025 View Certificate

Deep Learning MiniCamp [Arabic]

Udemy

Issued Apr 2025 View Certificate

Python for ML & Data Science Masterclass

Udemy

Issued Dec 2024 View Certificate

Python for Data Science, AI & Development

IBM (Coursera)

Issued Sep 2024 View Certificate

AI Python for Beginners

DeepLearning.AI

Issued Mar 2025 View Certificate

Introduction to Generative AI Learning Path

Google Cloud

In Progress View Course

Get In Touch

I'm actively seeking internship opportunities and collaborations in AI/ML and Data Science. Whether you have a project idea or just want to connect, feel free to reach out!

Email

moustafa.mh.mohamed@gmail.com