Skills And Expertise

Technical Skills

Programming Languages

  • Python: Experienced in machine learning, deep learning, and software development
  • C++: Intermediate knowledge with experience in algorithm implementation and briefly for computer vision

Machine Learning & Deep Learning

  • Frameworks: PyTorch, TensorFlow, Keras
  • Libraries: NumPy, Pandas, Scikit-learn, Matplotlib
  • Specialized Tools (limited experience): NVIDIA TensorRT, CUDA

Software Engineering Skills

  • Open-source library development
  • Agile methodologies (Kanban)
  • Asynchronous and multiprocessing programming
  • Git version control
  • API integration and microservices architecture

Open Source Projects & Contributions

ML Libraries

  • Developed and deployed multiple Python libraries on PyPI, including TransformerX, Emgraph, and Bigraph
  • Implemented modular, object-oriented architectures with comprehensive documentation and testing
  • Actively engaged in open-source community through code reviews and collaborative development

TASE: Telegram Music Search Engine

  • Built a scalable music search engine using Python
  • Integrated technologies: Elasticsearch, Pyrogram, ArangoDB, RabbitMQ, Celery
  • Implemented fault-tolerant microservices architecture

Technical Writing

Languages & Personal Interests

Languages: Fluent in English, Kurdish, and Persian.

Interests: I enjoy staying up-to-date on the latest developments in the field of artificial intelligence, and regularly read outstanding papers in the field. In my free time, I enjoy strolling around the city or hiking.

Other Experiences in Research Paper Re-implementations During Projects or Courses

Computer Vision

  • Object Detection and Recognition
    • YOLO v3 for car detection and object localization (CVPR 2015)
    • ResNet50 for hand sign recognition (CVPR 2015)
    • Transfer learning and fine-tuning pretrained models for car recognition
  • Face Recognition
    • Inception network and FaceNet implementation
    • Face recognition embedding based on Schroff et al. (2015)
  • Generative Models
    • Neural style transfer (Gatys et al., 2015)
    • Volcanoes classification on Venus dataset
    • Fruit image generation using Variational Autoencoders
    • Partial implementation of Make-A-Video: Text-to-Video Generation (2022)

Natural Language Processing

  • Machine Translation
    • Neural machine translation using attention mechanisms
    • Transformer model implementation for French-to-English translation (using TransformerX library)
  • Text Generation and Classification
    • Fine-tuning TF-Hub pretrained models for text classification
    • Character-level text generation using RNNs
    • Word-level text generation (Shakespeare-style poem generation)
    • Sentiment classifier with emoji recommendation using LSTM and GloVe embeddings
    • Tweet emotion recognition
  • Embedding and Representation Learning
    • Sequence models with various embedding layers
    • Word vector representations (Word2Vec, GloVe)
    • Sequence models using attention mechanisms

Speech Processing

  • Jazz solo improvisation using LSTM on music corpus
  • Deep learning for trigger word detection