Ahmad KHEDER

I build   |

Hey — I'm a Computer Engineer specializing in AI & ML who turns data into intelligence. Deep learning, transformers, computer vision, time-series — I build the models that make things actually think.

Scroll

About Me

Education

B.Sc. in Computer Engineering

Graduated from Hitit University, Türkiye

Languages

Arabic Native
English Professional
Turkish Intermediate

Interests

Artificial Intelligence Autonomous Systems Deep Learning Research NLP Systems

Computer Engineer passionate about building intelligent systems from scratch — from training deep learning models on raw data to deploying them into real-world production environments. Skilled in ML algorithms and classification systems, with hands-on experience in transformer architectures, computer vision, and time-series forecasting.

Technical Skills

Programming

Python
C++
C

Deep Learning

BERT DistilBERT CNN LSTM Transformers Transfer Learning EfficientNetB0 ML Algorithms Classification

Frameworks

PyTorch TensorFlow Keras HuggingFace Scikit-learn Flask

Optimization

AdamW LR Scheduling Gradient Clipping EarlyStopping Regularization

Web & Deployment

REST APIs Flask Backend API Development CPanel Web Hosting

Tools

Pandas NumPy Git Linux Jupyter GPU Training

Experience

AI & Software Developer

Independent / Self-Directed

2024 – Present
  • Designed and trained end-to-end deep learning systems covering NLP, Computer Vision, and Time-Series domains.
  • Fine-tuned BERT and DistilBERT transformer models using PyTorch and HuggingFace Transformers library.
  • Implemented transfer learning pipelines with EfficientNetB0 and custom TensorFlow/Keras architectures.
  • Integrated trained ML models into Flask REST APIs for production-style deployment and real-time inference.
PythonHuggingFaceTensorFlowFlaskBERT

Projects

01

Weather AI

Multi-City Forecasting System

Built a comprehensive weather forecasting system by training 40 trained .pkl models, each serving real-time weather predictions. Deployed through a Flask REST API with a modern frontend interface.

LSTMRegressionFlask APIPythonNumPy
40.pkl Trained Models
02

NLP Classification Systems

BERT / DistilBERT Fine-Tuning

Fine-tuned state-of-the-art transformer models on large-scale text classification datasets. Applied advanced techniques including AdamW optimizer, learning rate schedulers, gradient clipping, and comprehensive evaluation metrics.

BERTDistilBERTHuggingFaceTorchAdamW
SOTATransformer Models
03

Plant Disease Classification

EfficientNetB0 + CNN Transfer Learning

Developed a plant disease detection system using EfficientNetB0 CNN with transfer learning. Applied data augmentation, callback-based optimization (EarlyStopping, ReduceLR), and regularization to achieve high validation accuracy.

EfficientNetB0CNNTensorFlowKerasTransfer Learning
HighValidation Accuracy

Get In Touch

For professional opportunities, collaborations, and technical discussions.

contact@ahmadkheder.me