Contact
[email protected]
Ilya Shigabeev
Machine Learning Engineer with a focus on Deep Learning for
www.linkedin.com/in/shigabeev Speech.
(LinkedIn) San Jose, California, United States
Top Skills Summary
Stable Diffusion
Diffusion Tensor Imaging
I'm a strong Machine Learning software engineer with 5 years
Data Acquisition of experience in Natural Language, Speech Processing, Deep
Learning, and development of applied machine learned systems. I'm
Languages passionate about experimenting with algorithms and applying data
Russian (Native or Bilingual) science to build groundbreaking products. I've successfully coached
English (Native or Bilingual) and led teams through research and implementation projects
working closely with product managers and senior commercial
Certifications stakeholders. I enjoy working in high performance teams in a fast-
CCNA paced environment and am looking to apply my skills at scale in a
Startup School Certificate of leading tech company that is transforming how people interact with
Completion technology.
TOEFL Score 90
IELTS Score 6.5
Experience
Honors-Awards
1st Place on Startup weekend Cantina Labs
1 year 3 months
Bronze on Kaggle's Ship Detection
Championship Data acquisition, Image Generation Team
1st place in Makeathon May 2024 - Present (10 months)
Santa Clara, California, United States
Publications Invented a handful of tools to acquire data in high quantities for Image
Biochemical markers of ischemia, generators
heart failure and myocardial
deformation of children with
Machine Learning Researcher, Voice Team
hypertrophic cardiomyopathy
December 2023 - May 2024 (6 months)
Matching free-typed addresses
Santa Clara, California, United States
Predicting interests of Twitter users
using Machine Learning Training models for voice cloning
The ID R&D System Description for
Short-duration Speaker Verification Lang Swap App
Challenge 2021
Founder
December 2022 - Present (2 years 3 months)
Remote
Making video translations available to wide spectrum of content creators.
Page 1 of 5
KnowCorp
Machine Learning Engineer: Voice Cloning
May 2022 - December 2022 (8 months)
Los Angeles, California, United States
Created a text-to-speech synthesis system that only needs 10-second audio
from a source speaker to copy voice
Found multiple sources for audio data with speakers with rich pronunciation,
inflections, and intonations, which the model inherited
Participated in numerous other ad-hoc projects, such as chatbot finetuning
ID R&D
Sr. Machine Learning Researcher
December 2020 - May 2022 (1 year 6 months)
St Petersburg, St Petersburg City, Russia
We are working on creating the most secure and robust voice biometry system
in the world.
SberDevices
Machine Learning Researcher: Text to Speech
February 2020 - May 2020 (4 months)
Moscow
Successfully trained multiple speech synthesis models, including Tacotron
2, FastSpeech, and LPCNet, as part of a research project to build the first
intelligent voice assistant for the largest bank in Russia, Central and Eastern
Europe.
Designed and launched distributed training for a MelGAN model on the world's
7th most powerful supercomputer built for tasks in AI.
OZON.ru
Data Analyst
October 2019 - January 2020 (4 months)
Moscow, Russian Federation
Increased percentage of packages company delivers on time (credits here to
my manager, Lida, for her excellent strategy).
Implemented a system that daily processes and resolves issues with delayed
items.
Added detailed info about status of delayed items for call-center.
Taught Machine learning course to my department.
Page 2 of 5
National University of Science and Technology "MISIS" (Moscow
Institute of Steel and Alloys)
Teaching Assistant
September 2019 - January 2020 (5 months)
Moscow, Moscow City, Russia
Assistant on course of Machine Learning
Cisco
Data Scientist
August 2018 - July 2019 (1 year)
San Francisco Bay Area
Led the design, development and pilot of a word2vec-based machine
learned system that automated customer support using user-written textual
descriptions, achieving over 30% ticket response rate during pilot phase
across 200+ customers.
Researched and evaluated state of the art approaches to text vectorisation,
such as ELMO, FastText and BiLSTM.
Built a containerised, scalable and secure web service to label over 20k data
points, which was used to develop a machine learned ranking that reduced the
error rate of automated customer support ticket responses by 50%.
OZON.ru
Business Data Analyst
March 2018 - August 2018 (6 months)
Moscow, Russian Federation
Developed software to collate, parse, and report customer reviews from
multiple public review forums to Facebook Workplace in real-time.
Built a new executive reporting framework in PowerBI and implemented spatial
analysis to support the CEO-led revenue growth and customer experience
improvement programme.
Set up operational reporting and carried out ad hoc analysis for the logistics
department that unlocked an increase in on-time deivery from 50% to over
80% over 3 months.
National University of Science and Technology "MISIS"
Teaching Assistant
February 2018 - June 2018 (5 months)
Moscow, Russian Federation
Teaching Assistant on course "Artificial Intelligence: Principles and
Techniques"
Page 3 of 5
Promsvyazbank
Data Scientist
December 2017 - March 2018 (4 months)
Moscow, Russian Federation
Improved the regression model for predicting corporate client default risks by
enahncing data quality, which led to a 50% error rate reduction.
Created an address validation tool using elastic search achieving 92%
accuracy even when using rough free text data.
Argumenti nedeli
Press Photographer
April 2014 - October 2016 (2 years 7 months)
Moscow, Russian Federation
Took pictures of the Moscow's most beautiful celebrations and the most
engaging political protests and events. Was in tight contact with Moscow's
political activists.
FabLab 77
CAD Designer
January 2014 - October 2014 (10 months)
Moscow
Fabrication Laboratory is a place where one can make (almost) everything.
I am a recognised member of this lab. In FabLab I was designing furniture,
miniature robots and got my first experience with 3D printing. Also I had lots
of hands-on work with CAD, CAM, 3D printer slicing software and electronics
debugging.
Education
National University of Science and Technology "MISIS" (Moscow
Institute of Steel and Alloys)
Master's degree, Computer Science · (2017 - 2020)
National University of Science and Technology "MISIS" (Moscow
Institute of Steel and Alloys)
Bachelor of Science (BS), Computer Science · (2013 - 2017)
Stanford University
ICME
Page 4 of 5
Harvard University
Computer Science · (2016)
Higher School of Economics
Machine Learning · (2016)
Page 5 of 5
