Sign in Subscribe

Gor Matevosyan

A Bit About Me

As a data scientist based in Yerevan, Armenia, my focus is on projects related to Digital Signal Processing, Natural Language Processing, and Computer Vision. I have extensive experience in AI/ML research and data science, having worked as a Lead AI/ML Research Scientist at Podcastle AI and a Data Scientist at Teamable, where I led and contributed to complex projects involving Text-to-Speech, Speech Enhancement, Voice Cloning, job-candidate matching, and Computer Vision. Additionally, I have experience teaching Deep Learning and Neural Networks as a part-time Lecturer at Russian-Armenian University.

Along with my professional interests, I enjoy playing basketball, photography, painting, and travelling in my free time.

Experience

Lead AI/ML Research Scientist at Saima, Mar 2023 - present

  • Developed the main algorithm which stands behind Saima technology.

Lead AI/ML Research Scientist at Podcastle.ai, Jun 2020 - Mar 2023

  • Led the decision-making process for a complex project involving Text-to-Speech, Speech Enhancement, and Voice Cloning, collaborating with cross-functional teams.
  • Successfully built, trained, and improved high-quality Text-to-Speech models for production environments.
  • Developed custom Speech Enhancement models using state-of-the-art techniques that surpass the performance of existing open-source models.
  • Implemented efficient audio data pipelines for data processing and augmentation utilizing advanced Digital Signal Processing (DSP) algorithms.

Data Scientist at Teamable, Jan 2018 - Jul 2020

  • Developed innovative ideas for creating a job-candidate matching system by training customized Natural Language Processing (NLP) models. Significantly enhanced the matching accuracy of the top 10 results, achieving a score of 41%, which represents a substantial improvement from the previous 4%.
  • Developed and trained Computer Vision models for recognizing human diversity attributes such as age, gender, and ethnicity.

Lecturer at Russian-Armenian University, Sep 2020 - Jan 2023

  • Designed course plans for Deep Learning and Neural Networks.
  • Delivered lectures on Deep Learning and Neural Networks to Master's students in Computer Science.

Software Engineer, VR at Rodin.io, May 2016 - Dec 2017

  • Created an open-source JavaScript-based VR library using WebGL as a Virtual Reality Developer at Rodin Inc.
  • Created browser-side ES6 transpiler to analyse, transpile and run ES6 projects with imports and other futures directly in the browser faster 10 than existing solutions (babeljs).

Education

PhD Candidate at Russian-Armenian University, Spe 2018 - May 2024
Mathematical modelling with a focus on graph theory. Near-interval colouring of graphs.

Master of Data Science at Yerevan State University, Spe 2016 - May 2018
Thesis: Optical Character Recognition (OCR) for Armenian Language using Computer Vision and Deep Learning.

Bachelor of Informatics and Applied Mathematics at Russian-Armenian University, Spe 2012 - May 2016

Skills

Professional Skills: Machine Learning, Deep Learning, Digital Signal Processing (DSP), Natural Language Processing (NLP), Computer Vision (CV)
Programming Languages: Python (PyTorch, Tensorflow, NumPy, Pandas), JavaScript, C++, SQL (Postgres, MySQL, NoSQL)
Mathematical Knowledge: Strong understanding of Statistics and Probability, Calculus, Linear Algebra, Graph Theory, Computational Game Theory, and Optimization Theory, with the ability to apply them to solve complex problems in AI/ML and data science.
Languages: English (business advanced), Armenian (native), Russian (business advanced)