Switzerland

Data Scientist

Master of Engineering - MEng

Bachelor of Applied Science (B.A.Sc.)

High School Diploma

Pooya Mohammadi Kazaj

Senior Computer Vision Engineer

AI/ML

Data mining

Data visualisation

Python

Docker

GitHub

GitLab

The all-in-one AI library for Persian. (https://github.com/hezarai/hezar)

Zebracat is building AI video generation technology that will reinvent visual storytelling.
My responsibilities:
Storytelling PipeLine for video generation
Video Generation PipleLine using text-video models
Enhancing the microservice architecture of the video generation pipelines

Ariyadis is an Iranian company that carries out large projects for big companies, such as Irancell, the biggest telecommunication company in Iran.

Projects:

Development of a deep learning and machine vision application to extract essential information from identity cards presented by clients for identity authentication.
Development of a deep learning machine vision application for edge devices to detect defects in various fabrics.
Development of geocoding and reverse geocoding services using Iranian road data.
Development of a speech-to-text (ASR) service specifically designed for the Persian language.

AIMedic is an Iranian company that delivers AI-based medical solutions for various applications, including skin cancer, Corona Virus, Brain Stroke, etc. 
I personally was the head of the skin cancer detection section, in which we utilized cutting-edge classification and segmentation methods for determining Melanoma lesions from normal ones and extracting the lesions pixels from dermoscopy images, respectively.
I also was responsible for implementing MLOps methods for deploying, maintaining, monitoring, and automating different sections of deep-learning procedures.

I presented a brief history of NLP and CNN models. I talked about transformers, Bert, different convolutional architectures, object detection, and segmentation models.

Symo is an Iranian startup that presents an engine for visual search for fashion clothes.

    Dataset gathering using crawlers
    Object detection and segmentation for getting exact images
    Feature embedding extractor

I presented two successive deep-learning courses, namely, Deep Learning Fundamentals and Advanced Deep Learning at Khajeh Nasir University. The courses were comprised of the following topics:

 Logistic Regression, Fully connected neural networks, Metrics
 TensorFlow/Keras
 CNN, ResNet, UNet
 RNN, GRU, LSTM
 Auto-Encoder & GANs
 Deep Reinforcement Learning & DQN & A3C

Developed a desktop application for Stock Information Visualization, Storage & Strategy Tester for high-volume processes using Python (Multi-Processing, PyQT5), MySQL & Evolutionary Algorithms for testing different strategy methods and finding the best optimal solutions

    Developed a Stock Information Crawler using Python(Multi-Threading & Beautiful Soup). Using the above-mentioned techniques I created a crawler to get and save a huge amount of data in the MySQL database.

Developed a desktop application for Stock Information Storage & Tester
for high-volume processes using Python(Multi-Processing, PyQt5,
Cython) & MySQL.

Pooya Mohammadi Kazaj🇨🇭

About Pooya Mohammadi Kazaj

Expertise

Sessions with Pooya Mohammadi Kazaj

What mentees say