Natalia Tchemodanov

Data Scientist @Amazon | ex-Twitch | PhD

San Francisco Bay Area

About

Experience

  • Senior Data Scientist - Global Mile Team at Amazon
    Nov 2023 - Present · 2 yrs 8 mos

    Working on international supply chain problems.

  • Data Scientist at Twitch
    May 2017 - Jun 2023 · 6 yrs 2 mos

    I have all the memes

  • Data Science Fellow at Insight Data Science
    Jan 2017 - May 2017 · 5 mos

    •Built an image recognition web app for converting facial expressions to emoji. •Utilized Python, SciKitLearn, and openCV to build a predictive model for facial expression classification. Explored constraints for real-time emoji classification. • Developed web interface via Flask and Bootstrap, and deployed to AWS via nginx and gunicorn.

  • Graduate Researcher at Cognitive Neurophysiology Lab, UCLA Neurosurgery
    Sep 2011 - Oct 2016 · 5 yrs 2 mos

    • Utilized signal processing and spectral analysis methods to analyze neural recordings collected from human neurosurgery patients implanted with deep brain electrodes and performing memory tasks. • Discovered novel frequency-specific predictors of memory success. • Utilized MATLAB and UNITY3D to design and code 10+ experiments for human eyetracking and virtual environment research projects conducted with epilepsy patients implanted with electrodes. • Designed and coded closed-loop brain-machine interface experiment for real-time detection of signals associated with poor memory, and application of deep brain stimulation for memory modulation. Project was part of the effort to build a memory-restoring brain prosthesis (DARPA Restoring Active Memory program). Co-first authored findings in a paper currently under review at Nature Neuroscience. • Developed and deployed the signal processing and filtering portion of the lab’s data preprocessing pipeline. Assisted with design of pipeline architecture. Administered lab Subversion repo.

  • R&D Intern at HRL Laboratories, LLC
    Apr 2014 - Sep 2014 · 6 mos

    Developed unit testing infrastructure to maintain the MATLAB code pipeline. Streamlined and organized codebase for preprocessing MRI data.