Tetyana Loskutova, PhD, Developer in Grimsby, ON, Canada
Tetyana is available for hire
Hire Tetyana

Tetyana Loskutova, PhD

Verified Expert  in Engineering

Data Scientist and Developer

Location
Grimsby, ON, Canada
Toptal Member Since
June 17, 2022

Tetyana is a technology entrepreneur who strives to provide clients with end-to-end service when creating new software solutions or revamping old ones. Some of the projects she has completed include financial and accounting systems, ML-powered systems for NLP, forecasting, and anomaly detection. Tetyana has worked for clients in several countries and in various industries, such as energy, government, education, and biotechnology.

Portfolio

Mbaassist
Python 3, Deep Learning, Django, JavaScript, Scikit-learn, Databases, R...
Bonify, LLC
Data Science, APIs, Unstructured Data Analysis, Big Data...
Pivotal Consulting Group, LLC
Python, Machine Learning, Marketing Attribution, Business to Business (B2B)...

Experience

Availability

Part-time

Preferred Environment

Python 3, Amazon Web Services (AWS), Google Cloud Platform (GCP), Machine Learning, Python, Financial Modeling, Language Models, BERT, Document Parsing

The most amazing...

...project I've developed is an NLP system for extracting key information from regulatory reports of listed companies.

Work Experience

Director

2020 - PRESENT
Mbaassist
  • Developed a system for short-term market forecasting. The system tracks performance over time using the online dashboard on the company's website. The market forecasting system is also available through an API with free and paid options.
  • Became a top ML tutor in the Greater Toronto area.
  • Helped over 50 clients with small and medium projects for over a year.
Technologies: Python 3, Deep Learning, Django, JavaScript, Scikit-learn, Databases, R, TensorFlow, Reinforcement Learning, Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT), AWS Fargate, MongoDB, Linux, Data Science, Statistical Analysis, Decision Modeling, eCommerce, PyTorch, Machine Learning, NoSQL, Docker, Python, Amazon SageMaker, Non-fungible Tokens (NFT), Financial Data, Financial Modeling, Artificial Intelligence (AI), Text Analytics, Amazon Web Services (AWS), Jupyter, Amazon DynamoDB, Lambda Functions, Convolutional Neural Networks (CNN), Object Detection, Keras, Cloud Infrastructure, Large Language Models (LLMs), Business to Business (B2B), B2B, Marketing Mix Modeling, Marketing Attribution, BigQuery, Google Analytics, Full-stack, Software Architecture, Technical Leadership, Computer Vision, Social Media, Social Media APIs, Architecture, Language Models, PDF, BERT, Document Parsing, Deep Neural Networks, NLU, Test-driven Development (TDD), Market Research, Data Analytics, Data Analysis, Market Research & Analysis, REST APIs, Leadership, CTO, Google Cloud Datastore, Google Cloud, OCR, AWS Lambda, Generative Artificial Intelligence (GenAI), Dashboards, Generative AI

Data Scientist

2024 - 2024
Bonify, LLC
  • Built a topic labeling and classification tool for customer support tickets. Used a combination of classification, large-language, and statistical models to achieve the best cost-benefit ratio.
  • Developed a web app to review and analyze topics and subtopics found in the user data. Deployed the app as a service on AWS.
  • Built a tool to collect the data from an API and clean and organize the data according to dates, data attributes, and NLP-derived topics.
Technologies: Data Science, APIs, Unstructured Data Analysis, Big Data, Amazon Web Services (AWS), MySQL, User Journeys, Customer Journeys, Amazon API Gateway, Mistral AI, Gemini API, Natural Language Processing (NLP), Topic Modeling, Semantic similarity, Streamlit, Visualization, Hugging Face, Generative AI

Data Scientist

2023 - 2023
Pivotal Consulting Group, LLC
  • Developed several multichannel attribution models for marketing activities of PCG clients. The models were based on the Markov-Chain process, first and last touch attribution, and linear attribution modeling.
  • Developed an API application for managing PCG clients, auto-generating pixels, enabling and disabling client websites, and channel tracking. My role included the design, architecture, and full implementation of a Django app on Cloud Run.
  • Managed a Google Cloud platform and the database collecting data for client modeling.
  • Assisted the front-end developer with understanding the requirements and front-end implementation.
  • Automated the model calculation on a daily basis to improve the API's responsiveness.
  • Implemented queues and schedules to manage customer click data asynchronously.
Technologies: Python, Machine Learning, Marketing Attribution, Business to Business (B2B), B2B, Data Science, Marketing Mix Modeling, Attribution Modeling, JavaScript, Google Analytics, Salesforce API, BigQuery

Machine Learning Manager

2019 - 2020
MultiChoice
  • Recruited and coached a team of ML engineers. I have built a team from scratch and completed a forecasting project, content recommendation system, and fraud detection system.
  • Implemented an agile development process within my team, which served as an example for other teams in our department.
  • Established stakeholder cooperation across the company. I have led data science and ML programs across departments.
Technologies: Python 3, Scikit-learn, Django, GraphQL, Azure ML Studio, Databricks, Spark, Databases, Reinforcement Learning, Linux, Data Science, Statistical Analysis, Decision Modeling, Machine Learning, Docker, Python, Amazon SageMaker, Financial Data, Financial Modeling, Artificial Intelligence (AI), Text Analytics, Amazon Web Services (AWS), Jupyter, Convolutional Neural Networks (CNN), Keras, Cloud Infrastructure, Azure, Marketing Mix Modeling, Marketing Attribution, Full-stack, Software Architecture, Technical Leadership, Social Media, Social Media APIs, Architecture, Language Models, PDF, Document Parsing, Deep Neural Networks, NLU, Test-driven Development (TDD), Data Analytics, Data Analysis, Kubernetes, REST APIs, Leadership, Chatbots, AWS Lambda, Dashboards

Chief Data Scientist | Founder

2018 - 2019
Futuresgroup NL
  • Designed and implemented a system for tracking international trade. The system processed and reconciled a large volume of data to analyze and forecast international trade and changes in political situations.
  • Researched, designed, and developed an international investment risk and opportunity analysis dashboard. The system was based on structured and unstructured data, including news, collected from various sources.
  • Implemented an agile development process in my team.
Technologies: Python 3, SpaCy, Dash, Tableau, MongoDB, Linux, Databases, Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT), Django, Scikit-learn, Data Science, Statistical Analysis, Decision Modeling, Machine Learning, NoSQL, Docker, Python, Modeling, Financial Data, Financial Modeling, Artificial Intelligence (AI), Text Analytics, Amazon Web Services (AWS), Jupyter, Cloud Infrastructure, Business to Business (B2B), B2B, Full-stack, Software Architecture, Technical Leadership, Social Media, Social Media APIs, Architecture, Language Models, PDF, Computer Vision, Document Parsing, NLU, Test-driven Development (TDD), Data Analytics, Data Analysis, REST APIs, Leadership, CTO, Dashboards

Consultant

2017 - 2018
Control Risks
  • Used anomaly detection algorithms to detect transactional data. These data helped the conviction of members of a credit fraud syndicate in a bank.
  • Detected insurance fraud syndicates using machine learning algorithms.
  • Worked on fraud detection projects with teams in the Middle East, South Africa, and the USA, preprocessing large volumes of data for analysis.
  • Developed sales lead tracking and management web system.
Technologies: Python 3, SAP, Tableau, QlikView, Databases, Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT), Django, Scikit-learn, Data Science, Statistical Analysis, Decision Modeling, Machine Learning, Python, Modeling, Financial Data, Financial Modeling, Artificial Intelligence (AI), Text Analytics, Amazon Web Services (AWS), Lambda Functions, Keras, Cloud Infrastructure, Business to Business (B2B), B2B, Full-stack, Software Architecture, Technical Leadership, Social Media, Social Media APIs, Architecture, Language Models, PDF, Document Parsing, NLU, Test-driven Development (TDD), Data Analytics, Data Analysis, REST APIs, Leadership, Dashboards

IT Consultant

2007 - 2017
EOH Holdings
  • Built several ML models for energy demand forecasting.
  • Developed an SMS-based scheduler and notification system for power station maintenance.
  • Used data science tools to develop a municipality revenue improvement system.
  • Created the design and architecture of web-based climate change monitoring and evaluation system.
Technologies: Python 3, .NET, Java, Oracle, R, SQL, Databases, Linux, Data Science, Statistical Analysis, Decision Modeling, NoSQL, Python, Financial Data, Financial Modeling, Business to Business (B2B), B2B, Full-stack, Software Architecture, Technical Leadership, Architecture, Document Parsing, Test-driven Development (TDD), Data Analytics, Data Analysis, REST APIs, Leadership, Apache Kafka, Dashboards

Detection of Brain Tumors On MRI

https://drive.google.com/file/d/1u6tE74yz2sFzVKVc1EOvbek6aicasWGH/view?usp=sharing
I worked with a researcher to create a deep learning model that detected the shape of a brain tumor, if present, on MRI scans.

The input was taken from T1 and T2 scans saved in *.dcm files, and the output was presented in 3D Slicer software.

VoiceAPI

https://github.com/tetyanaloskutova/VoiceAPI
An API to generate voice recordings from a voice sample and text.

The API is deployable on AWS Fargate and uses the Tacotron PyTorch model for training and inference. Apart from voice generation, the API provides an option to fine-tune the model on new samples.

Day-Ahead Market Price Forecast Based on Media Trends

https://rapidapi.com/remote-works-default/api/media-obsessed-market/
This project uses historical price moves and sentiment changes in the media to predict price moves for a range of stocks for the following day. The predictions are based on a number of ML models (linear models, ensemble models, and LSTM). The models are evaluated for the previous day, and the best-performing model is chosen for the next day. The project's historical performance is tracked and is available as an API endpoint and an online dashboard:
https://mbaassiste.herokuapp.com/
2020 - 2021

Master's Degree in Physics

McMaster University - Hamilton, ON, Canada

2014 - 2017

Ph.D. in Business

University of the Witwatersrand - Johannesburg, South Africa

NOVEMBER 2020 - PRESENT

Les Fondamentaux de la Négociation

ESSEC Business SchoOl Through Coursera.

MARCH 2015 - PRESENT

Economics of Competing Energy Sources

Stanford University

APRIL 2013 - PRESENT

Certified Wireless Engineer

Duxbury

DECEMBER 2012 - PRESENT

Certified Energy Manager

Association of Energy Engineers

MAY 2011 - PRESENT

Oracle SQL

Oracle

Libraries/APIs

TensorFlow, Scikit-learn, Social Media APIs, REST APIs, PyTorch, Keras, SpaCy, Salesforce API

Tools

Amazon SageMaker, Jupyter, BigQuery, Google Analytics, AWS Fargate, Azure ML Studio, Tableau

Frameworks

Django, Spark, .NET, Streamlit

Languages

SQL, Python 3, Python, R, JavaScript, GraphQL, Java

Paradigms

Data Science, B2B, Test-driven Development (TDD)

Platforms

Oracle, Google Cloud Platform (GCP), Amazon Web Services (AWS), Azure, AWS Lambda, Docker, Kubernetes, Databricks, Linux, QlikView, Heroku, Apache Kafka

Storage

Databases, NoSQL, MongoDB, Amazon DynamoDB, Google Cloud Datastore, Google Cloud, Amazon S3 (AWS S3), BigQ, MySQL

Industry Expertise

Social Media

Other

Natural Language Processing (NLP), Deep Learning, Statistical Analysis, Decision Modeling, Machine Learning, Financial Data, Financial Modeling, Artificial Intelligence (AI), Text Analytics, Lambda Functions, Cloud Infrastructure, Large Language Models (LLMs), Business to Business (B2B), Full-stack, Software Architecture, Technical Leadership, Computer Vision, Architecture, Language Models, PDF, BERT, Document Parsing, Deep Neural Networks, NLU, Data Analytics, Data Analysis, Leadership, Generative Pre-trained Transformers (GPT), Dashboards, eCommerce, Modeling, Convolutional Neural Networks (CNN), Object Detection, Marketing Mix Modeling, Marketing Attribution, Market Research, Market Research & Analysis, CTO, Chatbots, OCR, Marketing Research & Analysis, Generative AI, Reinforcement Learning, Biostatistics, Numerical Methods, Statistics, Tax Accounting, Applied Physics, Negotiation, Wireless, Dash, SAP, Non-fungible Tokens (NFT), Regression, APIs, Attribution Modeling, Generative Artificial Intelligence (GenAI), Unstructured Data Analysis, Big Data, User Journeys, Customer Journeys, Amazon API Gateway, Mistral AI, Gemini API, Topic Modeling, Semantic similarity, Visualization, Hugging Face

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring