ooleksyuk.github.io

Olga Oleksyuk

mail@olga-v.com

Summary

I’m interested in machine perception in robotics.

I’m eager to contribute to the next revolution in mobility and what might be the most profoundly impactful technological advancement: self-driving cars. With multiple years of experience in software development, as well as team leadership, and demonstrable communication skills, I have much to offer and embrace the challenge in pivoting my career toward artificial intelligence and robotics.

I think allowing machines to see will unlock the revolution in AI. I want to be a part of that revolution.

Technologies: C++, OpenCV, Eigen, Python, Tensorflow, Keras.

Experience

• Self-education in self-driving cars and machine perception

  February 2017 – March 2018 (13 months)

  • Self-Driving Car Engineer Nanodegree [Udacity]
    • Traffic Sign Classifier (LeNet Network)
    • Lane Finding (Camera Calibration, Perspective transform, Polynomial fit)
    • Vehicle Detection and Tracking (HOG, Linear SVM classifier, sliding-window, heatmaps)
    • Mapping and Localization (Particle Filter, Extended, and Unscented Kalman Filters)
    • Throttling and Steering (Model Predictive Control, Behavioral Cloning, PID Controller)
    • Path Planning (A* Algorithm, Sensor Fusion, Prediction, Behaviour Planning, Trajectory Generation)
    • Scene Understanding (Semantic Segmentation, FCN, VGG16, KITI, CityScapes, Upsampling, Cross-Entropy Loss)
    • Programming a Real Self-Driving Car (ROS, U-Net, Sequential, PID Controller, Trajectory Planner, Waypoint Follower, Stability Controller, Gain Controller)
  • Convolutional Neural Networks [Coursera]
  • Improving Deep Neural Networks: Hyperparameter tuning, Regularization, and Optimization [Coursera]
  • Structuring Machine Learning [Coursera]
  • Neural Networks and Deep Learning [Coursera]
  • Machine Learning [Coursera]

• Software Engineer at ezCater

  January 2016 - Present (2+ years), Boston

  Back-end development using PostgreSQL, Sidekiq, Amazon Kinesis Firehose, Amazon RedShift, Ruby on Rails.

  Built a high-volume data transformation and synchronization tool.

  Lead integration with a third-party restaurant management system. Identified and automated a number of steps in the ordering processing.

  Developed the system for collecting the detailed logs of users’ interactions with the website. Built tools for the transformation of the collected data into the format which could be analyzed by machine learning tools.

• Maternity leave

  December 2011 - December 2015

• Full Stack Developer, Contractor

  2008 – 2011 (3 years), Ukraine

  Developed back-end and front-end for business applications. Mostly used Python and C#/.NET, and business applications using Node.js, Express, and MongoDB, PHP/MySQL, HTML, CSS, JavaScript.

Education

• Taras Shevchenko National University of Kyiv
• Master’s degree in, Applied Mathematics / Computer Science, 2006 - 2008
• Taras Shevchenko National University of Kyiv
• Bachelor’s degree in, Applied Mathematics / Computer Science, 2002 - 2006