AI Engineer  ·  Machine Learning  ·  Software

Hi, I'm Ahmed —
I build things that
learn from data

I'm a

Designing, training, and deploying intelligent machine-learning systems and modern, human-centered web solutions.

View Work Contact Me
  • Python
  • PyTorch
  • TensorFlow
  • FastAPI
  • Computer Vision

01 About

From raw data to deployed, working product — I care about the whole arc.

I'm a passionate AI & Software Engineer focused on building scalable machine-learning pipelines, deep-learning models, and interactive software applications. I turn complex data into actionable, intelligent solutions — and I obsess over the details that make them usable in the real world.

From feature engineering to model evaluation to a clean API your product team can actually ship with — I enjoy being hands-on at every stage.

Projects delivered
15+
Average model accuracy
98%
Years of experience
3+

02 Education

My academic foundation

B.Sc. Candidate in Information Systems (IS)

Faculty of Computers and Artificial Intelligence, Benha University (BFCAI)

2023 – 2027 (Expected)

  • Database Systems
  • System Analysis & Design
  • Machine Learning
  • Data Analytics
  • Software Engineering
3.0/ 4.0 GPA

03 Skills

What I work with every day

Machine Learning & AI

Training robust models and pushing them to production with a measurable, testable approach.

  • PyTorch
  • TensorFlow
  • Scikit-Learn
  • OpenCV
  • NLP
  • Computer Vision

Data & Engineering

Cleaning, shaping, and extracting value from data before a single model is ever trained.

  • Python
  • SQL
  • Pandas
  • NumPy
  • Data Mining
  • Feature Engineering

Web & API Development

Wrapping intelligence in fast, well-documented interfaces that teams love to integrate.

  • HTML5
  • CSS3
  • JavaScript
  • FastAPI
  • Streamlit

04 Projects

Selected work

House Features Inputs Figure 1: Application UI & Primary Inputs
Secondary Features Inputs Figure 2: Secondary House Properties
House Price Prediction Result Figure 3: Estimated Price Output
Actual vs Predicted Prices Plot Figure 4: Actual vs Predicted Prices Plot
Model Evaluation Score Figure 5: Gradient Boosting Regressor R² Score
1 / 5

House Price Prediction App

End-to-end ML web application using Gradient Boosting to estimate house market prices from structural and geographical features — R² score of 92.2%.

  • Python
  • Scikit-Learn
  • Gradient Boosting
  • Streamlit
  • Pandas
View on GitHub
Training Accuracy and Loss History Figure 1: EfficientNetB0 Training vs Validation Performance
Model Training Epochs Log Figure 2: Transfer Learning Training Logs (89% Val Accuracy)
FastAPI Swagger UI Endpoint Figure 3: FastAPI REST Endpoint (/predict) Interface
Inference Response - Corn Common Rust Figure 4: Real-time API Response (Corn Rust - 100% Confidence)
Test Sample Leaf - Corn Leaf Figure 5: Input Sample - Infected Corn Leaf
FastAPI Swagger UI Endpoint 2 Figure 6: API Testing with Apple Brown Spot Sample
Inference Response - Apple Brown Spot Figure 7: API Response Output (Apple Spot - 99.7% Confidence)
Test Sample Leaf - Apple Leaf Figure 8: Input Sample - Infected Apple Leaf
1 / 8

AI Plant Disease Classifier & API

Computer vision pipeline using EfficientNetB0 transfer learning to classify 70+ plant diseases at 89% accuracy, deployed behind a FastAPI backend.

  • TensorFlow
  • EfficientNetB0
  • FastAPI
  • Transfer Learning
  • Python
View on GitHub
YOLO Vehicle Detection Bounding Boxes Figure 1: Object Detection Output with Bounding Boxes & Confidence
FastAPI Vehicle Detection API Interface Figure 2: Vehicle Detection & Counting API (/detect) Endpoint
API Response JSON Vehicle Counts Figure 3: Real-time API Response JSON with Vehicle Class Breakdown
Tested Sample Traffic Image Figure 4: Tested Traffic Input Image
1 / 4

Vehicle Detection & Counting API

Real-time computer vision system powered by YOLO that detects and counts traffic (cars, buses, trucks, motorcycles) via a FastAPI REST endpoint.

  • YOLO
  • Computer Vision
  • FastAPI
  • OpenCV
  • Python
View on GitHub
EasyOCR Reader Initialization Figure 1: EasyOCR Pipeline Setup & GPU Acceleration
OCR Reader Code Loop Figure 2: Text Detection & Confidence Score Extractor
Extracted Text with Confidence Scores Figure 3: Raw Detected Text & Confidence Breakdown
FastAPI Receipt OCR Endpoint Figure 4: FastAPI OpenAPI (/extract) Endpoint Interface
Structured Receipt JSON Output Figure 5: Parsed JSON Output (Store, Total, Cash, Items)
1 / 5

Receipt OCR & Parsed Text API

Computer vision pipeline using EasyOCR and regex parsing to extract structured JSON — store name, items, total, cash received — served via FastAPI.

  • EasyOCR
  • Computer Vision
  • FastAPI
  • Regex Parsing
  • Python
View on GitHub

05 Contact

Let's build something together

Email ahmed365j@gmail.com LinkedIn Ahmed Nasr WhatsApp +20 112 616 9033 Telegram +20 112 616 9033

Send a quick message

Fill this in and it opens a direct WhatsApp chat with me — no sign-up, no waiting.