Professional skills:
More than 5 years in commercial development
Experience working in a team, incl. distributed
Experience managing a small team (3 people)
Mentoring experience
Teaching experience
Certificates and courses:
Samsung Russia Open Education course, Neural networks and natural language processing, Stepik, 2023
CertNexus course. Extract, Transform and Load Data, Coursera. 2022
HSE University course. Natural Language Processing, Coursera. 2021
Open Machine Learning Course, Open Data Science. 2019
Education:
-St. Petersburg State Electrotechnical University "LETI"
Faculty: Electrical Engineering and Automation
-St. Petersburg State Electrotechnical University "LETI"
Direction: System analysis, control and information processing
Frameworks and libraries:
DBMS:
|
IDE:
PyCharm
Spyder
Jupyter Notebook
MS Visual Studio
OS:
Windows
Ubuntu
Programming languages:
Python 3.7
MatLab
C
VCS:
GIT
DVC
Other:
JSON
Docker
Streamlit
MatLab
NLP (classification, clustering, NER, chat-bots, LLM etc.)
Prompt-engineering
PEFT
Airflow
MLFlow
PROJECTS
System for automatic quality assessment of the developed technical specifications
Project period | March 2023 - present |
Role | Data scientist, Developer |
Project’s area | IT services |
Description | A company that develops solutions to improve the process of producing technical documentation for large infrastructure projects needed to implement a system that would search for errors in documentation based on NLP approaches. |
Completed tasks | The tasks to be solved included:
To achieve this goal, the following tasks were solved: syntactic analysis of texts, selection and automatic generation of prompts for large language models (GPT, LLAMA2), training of language models |
Basic technologies | Python, PyTorch, huggingface-transformers, spacy, PEFT |
Research and development of generative language model
Project period | January 2023 – February 2023 |
Role | Data scientist, Developer |
Project’s area | IT services |
Description | For the industrial equipment marketplace, it was necessary to research and develop functionality for automatically filling product cards with images, descriptions, and a list of characteristics. The project solved the problem of generating a description of a product based on its name. |
Completed tasks | The completed tasks included:
|
Basic technologies | Python, PyTorch, huggingface-transformers |
Development of a chatbot for marketplace support service
Project period | January 2023 – March 2023 |
Role | Data scientist, Developer |
Project’s area | IT services |
Description | Chatbot for industrial goods marketplace. The chatbot's tasks include processing common questions to relieve the support team. |
Completed tasks | The completed tasks included:
|
Basic technologies | Python, PyTorch, huggingface-transformers, RASA |
Equipment shipment forecast
Project period | August 2022 – December 2022 |
Role | Data scientist, Developer, Analyst |
Project’s area | Logistics |
Description | For a large equipment supplier, it was necessary to predict the purchase of components to ensure timely replenishment of stocks in warehouses. The prediction was made for several periods in advance, from a month to a year. |
Completed tasks | The completed tasks included:
|
Basic technologies | Python, CatBoost, scikit-learn, PyTorch, matplotlib, seaborn |
Analysis of text sources
Project period | August 2021 – June 2022 |
Role | Data scientist, Developer |
Project’s area | IT Services |
Description | Development of a system that, based on open (news) and closed text sources, analyzes the geopolitical situation and generates some* recommendations. * under NDA |
Completed tasks | The tasks included all stages associated with processing text data. Stages: classification of input texts, extraction of named entities, parsing. The processing results were combined into an output JSON file that contained text in a structured form. The results obtained were processed by another team to solve the target problem. In addition, other studies related to natural language processing were carried out: text clustering, text ranking. |
Basic technologies | Python, PyTorch, HuggingFace-Transformers, SpaCy, Scikit-learn, airflow |
Implementation of IIoT infrastructure at hydroelectric power stations
Project period | August 2020 – May 2021 |
Role | Data scientist, Developer, Analyst |
Project’s area | Industry |
Description | Development and implementation of stationary devices for data exchange, cameras and wearable devices at hydroelectric power plants. The main tasks to be solved were monitoring workers at the station in order to quickly detect possible dangerous or emergency situations. |
Completed tasks | The completed tasks included:
|
Basic technologies | Python, CatBoost, scikit-learn, Docker, Cassandra, ClickHouse, matplotlib, seaborn |
Model of electric vehicle energy consumption at various loads
Project period | September 2019 – February 2020 |
Role | Developer |
Project’s area | Logistics |
Description | An external customer formulated the task of developing an electric vehicle model that should be combined with a customer-provided electric vehicle battery model. It was necessary to develop a model close to a real electric vehicle, which could be used to test the loads experienced by the electric vehicle battery in various operating modes. |
Completed tasks | The tasks included discussing the project with the team, searching and studying materials with suitable implementations of mathematical models of electric vehicles, implementing an electric motor in the MatLab environment and transferring the solution to Python. |
Basic technologies | MatLab, Python, SciPy, matplotlib |
Updating the technological process of the production line
Project period | September 2019 – November 2019 |
Role | Data scientist, Developer |
Project’s area | Industry |
Description | The polymer composite material production line underwent hardware and software upgrades to improve the characteristics of the output material. |
Completed tasks | The project consisted of a large number of tasks on which various teams worked. The following specific tasks were obtained and solved:
|
Basic technologies | Python, scikit-learn, sciPy, MatLab |
Honeycomb block cutting machine
Project period | August 2018 – August 2019 |
Role | Developer |
Project’s area | Industry |
Description | An industrial machine for cutting honeycomb blocks with high precision was developed |
Completed tasks | The tasks included:
|
Basic technologies | Omron PLC, Simple Scada |
GPS accuracy correction system
Project period | January 2018 – May 2019 |
Role | Developer |
Project’s area | Logistics |
Description | It was necessary to develop a system that improves the accuracy of a GPS system for a mobile robot based on other positioning methods (relative positioning using an odometry system, magnetometer and inertial sensors) by combining data from different devices. |
Completed tasks | Main goals:
|
Basic technologies | C programming language (software for microcontroller PIC24), MatLab |
Automation of country real estate
Project period | December 2016 – May 2018 |
Role | Developer |
Project’s area | IoT |
Description | For luxury suburban real estate, it was necessary, according to a pre-received plan, to develop the logic for the operation of smart home devices (lighting, motion sensors, control panels, etc.) under the control of a central controller according to the customer’s wishes |
Completed tasks | The tasks included:
|
Basic technologies | Programming language for Crestron controllers, Iridium Studio, JavaScript for interface development on iPad |
Получайте свежие вакансии первыми прямо на почту
Опередите остальных и первыми олткликнетесь на интересную работу
01.09.2019
25.08.2018
18.01.2019
Все страны
