Work Experience...

Technical Skills:

• Research Expertise: Machine Learning, Deep Learning, Computer Vision, Time-Series
• Programming: Java, C++, Python
• Deep Learning: PyTorch, Caffe, TensorFlow, Theano, Keras
• Database Management/Cloud: SQL, AWS, Docker Container

AI/ML Software Engineer                                                           Jun. 2021 - Present

Google 

AI Developer                                                                               Feb. 2019 - Jun. 2021

BMO AI Labs 

• Advancing software engineering and pushing the boundaries of machine learning. 

Software Engineer                                                                      Dec. 2017 - Dec. 2018

Ecopia AI

• Developed a two-staged algorithm for building footprint extraction by parsing satellite images, which resulted in having much more accuracy and speed in production (C++, Caffe, ResNet, CNN)
• Designed and implemented a deep algorithm for fine-grained road segmentation by parsing satellite images and road topology, which is used in production (C++, Caffe, Graph and Mathematical Modelling, ResNet)
• Developed C++ version of the whole MLFlow library, and added reproducibility (dockerization) [Code]

Co-Founder & CTO                                                                

ID Green Inc 
Funded by Entrepreneurship Hatchery, University of Toronto, with the goal of
 improving precision farming techniques using hyper-spectral imaging.

• Implemented a shrub segmentation algorithm (TensorFlow, U-net) 
• Designed and developed a method for rapid/accurate detection of nitrogen concentration from potato leaf hyper-spectral images for determining the health of the crops (TensorFlow, Stacked auto-encoders)

ML Researcher/Developer                                                         Sep. 2016 - Sep. 2017

Centre for Intelligent Mining Systems

• Designed and developed a novel ConvNet architecture (called Patched-CNN) for fast/robust shadow detection of natural scene images, which defeated state-of-the-art methods in speed (1.5 sec/image) and accuracy (%91) (Python, Theano), [Code], [Project Demo]
• Developed a robust visual place recognition algorithm (Caffe, Image matching, CNN) [Code]

Software Engineer Intern                                                          Sep. 2014 - Feb. 2015

ASR Gooyesh Pardaz Inc.

• Developed a fast vehicle license plate recognition software
  (LSTM Neural Networks, CURRENNT)

Teaching Assistant                                                                                                          

• University of Alberta:
  Algorithms I, Practical Programming Methodology (C,C++)
• University of Tehran:
  ACM-ICPC Programming, Intro to Java Programming, Advanced Programming (C++), Data Structures