Available for Remote Projects Worldwide

Building AI Solutions for Real-World Problems

I turn complex data into intelligent solutions, from predictive maintenance and forecasting to computer vision, NLP, and scientific machine learning.

Explore Projects → Discuss Your Project ↗
Salma Elgezawy
Machine LearningDeep LearningComputer VisionNLPTime-Series ForecastingScientific ML
About

About Me

I’m Salma Elgezawy, an AI/ML Engineer with a background in Artificial Intelligence and experience across machine learning, deep learning, computer vision, NLP, time-series forecasting, scientific machine learning, predictive maintenance, and Digital Twins.

My work combines data-driven modeling with experimentation and real-world engineering. During my research internship at Nile University’s Nanoelectronics Integrated Systems Center (NISC), I worked on chaotic systems, reservoir computing, and physics-informed machine learning, exploring how data-driven and physics-informed methods can improve forecasting of complex dynamical systems.

Across my projects, I focus on understanding the problem and the data first, then selecting, evaluating, and optimizing the approach that best fits the task.

ArabicEnglish
RESEARCH INTERNSHIP FEEDBACK

Research Summer Internship · NISC

Quality of Work5/5
Work Habits5/5
Interpersonal Skills5/5
Services

How I Can Help

AI and machine learning solutions designed around the problem, the data, and the decisions they need to support.

Machine Learning & Data Science

Turn structured data into predictive models and actionable insights through data preparation, feature engineering, model development, evaluation, and optimization.

Deep Learning & Computer Vision

Build deep learning solutions for image classification, object detection, semantic segmentation, and other visual recognition tasks.

Time-Series Forecasting

Analyze temporal data and develop forecasting models for trends, future values, and complex dynamical behavior.

NLP & Text Intelligence

Build NLP pipelines for text cleaning, representation, classification, and extracting useful structure from unstructured text.

Data Analysis & EDA

Explore datasets, uncover patterns, identify data-quality issues, engineer meaningful features, and prepare data for modeling.

Scientific & Physics-Informed ML

Develop machine learning approaches that combine data-driven modeling with physical knowledge for complex engineering and dynamical systems.

Portfolio

Selected Projects

Selected AI/ML projects across industrial systems, scientific forecasting, computer vision, NLP, and data-driven applications.

TwinSight predictive maintenance and Digital Twin project

TwinSight

An end-to-end predictive maintenance system combining sensing, signal processing, machine learning, RUL estimation, and a Unity-based Digital Twin.

Predictive MaintenanceIndustrial AIIoTDigital Twin
View Case Study →
Chaotic System Forecasting with Lorenz attractor visualization

Chaotic System Forecasting

A scientific ML project exploring long-horizon forecasting in chaotic dynamics using Echo State Networks and physics-informed reservoir computing.

Time SeriesReservoir ComputingPhysics-Informed MLOptimization
View Case Study →
Predictive Maintenance for Industrial Motors with vibration signal analysis and fault classification

AI-Powered Predictive Maintenance for Industrial Motors

A rotating-machinery fault-classification project using vibration signals, engineered spectral features, deep learning, and physics-informed modeling ideas.

Predictive MaintenanceSignal ProcessingPhysics-Informed MLFault Classification
View Case Study →
Oil Spill Semantic Segmentation

Oil Spill Semantic Segmentation

A multi-class semantic-segmentation project for aerial oil-spill imagery with class balancing, attention mechanisms, and a custom decoder.

Computer VisionSemantic SegmentationDeep LearningEnvironmental AI
View Case Study →
Resume Screening System with NLP pipeline for document classification

Resume Screening System

An NLP pipeline that cleans, represents, augments, and classifies resumes into job categories, with a Streamlit interface for document upload.

NLPText ClassificationMachine LearningData Analysis
View Case Study →
Weather Forecasting Project with Prophet and interactive visualization

Weather Forecasting

A practical forecasting project covering exploratory analysis, stationarity, seasonality, outlier handling, Prophet modeling, and interactive visualization.

Time SeriesForecastingData AnalysisProphet
View Case Study →
Skin Cancer Classification Project with multimodal deep learning combining dermatoscopic imagery and tabular metadata

Skin Cancer Classification

A research-oriented multimodal classification prototype combining dermatoscopic imagery with tabular metadata across seven lesion classes.

Deep LearningComputer VisionMultimodal MLMedical AI Research
View Case Study →
Approach

Why Work With Me

Problem-First Engineering

I start by understanding the problem and the data before choosing a model, with careful attention to preprocessing, EDA, feature engineering, validation, and modeling decisions.

Rigorous Experimentation

I compare approaches and strengthen ML pipelines through systematic evaluation, hyperparameter optimization, data augmentation, and class-balancing strategies.

From Models to Real Systems

My work goes beyond isolated models, combining AI with signal processing, IoT, Digital Twins, interactive applications, and domain-specific engineering when the problem requires it.

Skills

Technical Toolkit

Machine Learning & Data

PythonScikit-learnPandasNumPySQLXGBoostSVMDecision TreesEnsemble Learning

Deep Learning & Computer Vision

PyTorchTensorFlowCNNsOpenCVYOLOImage ClassificationSemantic SegmentationObject Detection

Forecasting & Scientific ML

Time-Series ForecastingFFT / Signal ProcessingPINNsReservoir ComputingESNPI-ESNOptuna

Tools, Integration & Optimization

StreamlitFlaskFastAPIMQTTUnityGrid SearchRandom SearchData Augmentation
Recognition

Achievements

Huawei ICT Competition · Cloud Track

Regional & National Recognition

A connected competition journey across the national and North Africa regional phases.

10thPlace, North Africa
PlatinumMedal, National Phase
Top 12Cloud Track, Egypt

Have an AI/ML problem worth solving?

Explore the case studies, then discuss your project with me through GoLance.

Discuss Your Project ↗