Hi, I’m Ali Mohamed Hashish, a machine learning engineer focused on NLP, computer vision, and MLOps. I enjoy turning data into deployable AI solutions and collaborating with cross-functional teams to ship real-world AI products. I have hands-on experience building end-to-end ML pipelines using PyTorch, TensorFlow, Hugging Face, FastAPI, Docker, and AWS ECS, and I enjoy explaining complex concepts to trainees and stakeholders. Hi, I’m also an AI instructor with a knack for clear communication and teamwork. I’ve guided over 30 trainees and mentored ML projects, and I love helping others grow while delivering practical AI solutions.

Ali Mohamed Hashish

Hi, I’m Ali Mohamed Hashish, a machine learning engineer focused on NLP, computer vision, and MLOps. I enjoy turning data into deployable AI solutions and collaborating with cross-functional teams to ship real-world AI products. I have hands-on experience building end-to-end ML pipelines using PyTorch, TensorFlow, Hugging Face, FastAPI, Docker, and AWS ECS, and I enjoy explaining complex concepts to trainees and stakeholders. Hi, I’m also an AI instructor with a knack for clear communication and teamwork. I’ve guided over 30 trainees and mentored ML projects, and I love helping others grow while delivering practical AI solutions.

Available to hire

Hi, I’m Ali Mohamed Hashish, a machine learning engineer focused on NLP, computer vision, and MLOps. I enjoy turning data into deployable AI solutions and collaborating with cross-functional teams to ship real-world AI products. I have hands-on experience building end-to-end ML pipelines using PyTorch, TensorFlow, Hugging Face, FastAPI, Docker, and AWS ECS, and I enjoy explaining complex concepts to trainees and stakeholders.

Hi, I’m also an AI instructor with a knack for clear communication and teamwork. I’ve guided over 30 trainees and mentored ML projects, and I love helping others grow while delivering practical AI solutions.

See more

Experience Level

Expert
Expert
Expert
Expert
Expert
Expert
Expert
Intermediate
Intermediate
Intermediate
Intermediate
Intermediate
See more

Language

Arabic
Fluent
English
Advanced

Work Experience

AI Instructor at ISET Academy
March 1, 2024 - September 1, 2025
Guided over 30 trainees in building ML and DL projects (supervised/unsupervised learning, classification, regression) with TensorFlow/Scikit-learn, delivering hands-on instruction, project supervision, and curriculum support.
Deep Learning Intern at Orange Digital Hub Egypt
January 1, 2025 - March 1, 2025
Built image classification, object detection, and segmentation pipelines with CNNs and ResNet fine-tuning; applied data augmentation, image alignment, and GANs for synthetic data; fine-tuned transformer-based models (AraBERT, GPT) for NLP tasks; learned fundamentals of MLOps: deployment, monitoring, and scaling.

Education

Bachelor's degree in Computer Science at Tanta University, Faculty of Computer & Information
October 1, 2020 - July 1, 2024

Qualifications

Deep Learning & Machine Learning Specializations
January 1, 2025 - December 21, 2025
Deep Learning for Computer Vision
January 1, 2025 - December 21, 2025
Associate Data Scientist
January 1, 2025 - December 21, 2025
Introduction to AI
January 1, 2024 - December 21, 2025
SQL
January 1, 2024 - December 21, 2025

Industry Experience

Software & Internet, Education, Computers & Electronics, Media & Entertainment, Professional Services
    paper Arabic Reviews Sentiment Analysis

    Arabic Reviews Sentiment Analysis is an AI‑driven NLP solution that classifies sentiment in Arabic text using deep learning and transformer‑based models. The system processes user reviews through text cleaning, tokenization, and model inference to determine positive, negative, or neutral sentiment with high accuracy.

    This project demonstrates proficiency in handling Arabic language challenges such as morphology, dialect variation, and contextual nuances. It includes a complete ML workflow with data preprocessing, model training, evaluation, and deployment‑ready inference components.

    Key Highlights

    • Sentiment classification tuned for Arabic language nuances
    • Transformer‑based modeling for improved contextual understanding
    • Optimized preprocessing to handle dialects, stopwords, and noise
    • Evaluation metrics with accuracy, precision, and recall
    • Clean, modular, and reproducible machine learning pipeline

    Business Value

    • Enables automated customer feedback analysis for product decisions
    • Supports brand monitoring and reputation insights
    • Useful for market research, support automation, and sentiment dashboards
    paper Offside Detection Pipeline: Soccer Match Analytics

    Offside Detection Pipeline is a computer vision system that identifies potential offsides from soccer match images by combining player detection, pose keypoint estimation, and geometric analysis. The pipeline extracts field lines, computes vanishing points, and visualizes offside lines based on player positions.

    The system integrates optimized segmentation, deep learning models (e.g., YOLOv8 for object detection and Keypoint R‑CNN for pose extraction), and clustering techniques to classify players and estimate accurate offside conditions. This project demonstrates practical expertise in sports analytics, computer vision, and model integration.

    Key Highlights

    • Player detection & classification into teams
    • Vanishing point estimation for accurate spatial context
    • Pose keypoint extraction for precise player positioning
    • Virtual offside line visualization for referee decision support
    • Modular ML pipeline suited for soccer analytics

    Business Value

    • Enables automated offside analysis for referee assistance and match review
    • Useful for sports performance analytics, broadcasting insights, and tactical evaluation
    • Scalable for integration into live match feeds or video analysis systems
    paper AREN-NMT: Arabic–English Neural Machine Translation

    AREN-NMT is an end-to-end Arabic–English AI translation system built using attention-based neural machine translation models. It delivers accurate and context-aware translations while addressing Arabic language complexity through optimized preprocessing and model design.

    The system covers the full machine learning lifecycle, including data preparation, model training, evaluation, and real-time inference. It is suitable for content localization, multilingual platforms, AI chatbots, and customer support automation.

    Key Highlights

    • Translation quality measured using BLEU score
    • Low-latency inference for real-time applications
    • Improved accuracy compared to baseline models
    • Scalable architecture supporting domain-specific fine-tuning
    • Clean and reproducible ML pipeline

    Business Value

    • Enables reliable Arabic–English localization for digital products
    • Supports AI-driven communication systems
    • Adaptable to industry domains such as medical, finance, and customer support