Vyacheslav Efimov
Machine Learning Engineer | Technical Writer

I am a passionate machine learning engineer with 4 years of experience in the intersection of software engineering and data science.I am constantly broadening my horizons by learning new technnologies and building projects!
Experience
Throughout my career I worked in different IT companies. This career experience played an integral role of making me better as an engineer.To find out detailed information on projects I was working on and skills I have obtained, check out the links below:
Education
I have studied at ITMO University which is known as one the best IT universities in the world of Computer Science.By studying an abundance of software engineering subjects and completing coding assignments, I have acquired necessary skills to start my career in development.
Portfolio
During my free time, I love building new projects! In my opinion, this is one of the best ways to build expertise. Check out what pet projects I have built by clicking the button below:
Additionally, I have my Medium blog where I publish Data Science articles on my favourite topics.Check out some of the best article series:
Contact me
If you have a question or would like to collaborate on a project, do not hesitate to reach me using the question form.
ยฉ Copyright. Vyacheslav Efimov.
Career ๐ผ
Title | Company | Location | Start date | End date |
---|---|---|---|---|
Machine Learning Engineer | Leveraize | Le Mans, France | Mar. 2024 | Present |
Machine Learning Engineer | Sberbank | Saint-Petersburg, Russia | Nov. 2021 | Sep. 2022 |
Machine Learning Engineer | Devim | Saint-Petersburg, Russia | Feb. 2021 | Nov. 2021 |
Intern Software Engineer | Element | Saint-Petersburg, Russia | June 2020 | Sep. 2020 |
Leveraize
Leveraize is a start-up specialized in the artificial intelligence development.
Implemented a framework for code generation using PuLP for solving linear optimization problems.
Created a full-stack application by developing frontend, backend, ML and DevOps modules for image segmentation with U-Net and deployed the whole application on AWS.
Developed YOLO and OCR models for detecting and recognizing medications in a live video stream and integrated the entire pipeline onto a Raspberry Pi.
Sberbank
Sberbank is the largest bank in Russia with an advanced AI ecosystem.
Improved ML metrics by developing cascade and blender ML models for predicting clientsโ payments based on transaction history for each of 35 product categories.
Implemented NLP model for multi-label user goals categorisation (โ100 classes) for products recommender system.
Developed Data Science tools in inner ML framework for the team.
Analysed and sped up join algorithms for receiving product merchants from 2GIS and merchants handbook data sources.
Performed adhoc analytics by developing ETL pipelines processing several TB of data.
Interacted with stakeholders and presented results throughout development process.
Devim
Devim is a fintech copmany developing credit scoring platforms.
Developed classification, regression, clustering and uplift ML models for credit scoring problems.
Reorganised dashboards in Grafana for more convenient analytics and usage.
Collected data from databases, external API sources.
Analysised A/B-tests of clientsโ statistics.
Element
Element specialises in development of smart city products.
Developed a REST API backend gateway for connecting 3G/4G/LTE devices to the IoT platform.
Increased testing coverage by writing series of unit tests.
Skills ๐ ๏ธ
Domain | Technologies |
---|---|
Programming Languages | Python, JavaScript, Java, C++ |
Machine Learning | Pandas, NumPy, Scikit-Learn, Seaborn, LightGBM, PuLP, TensorFlow, PyTorch |
Language Processing | HuggingFace, Faiss, Langchain, LLM |
Data | SQL, MongoDB, PySpark |
Frontend | HTML, CSS, React.js, Bootstrap, ChakraUI, Figma |
Backend | Node.js, Spring Boot, Flask, FastAPI, Streamlit |
DevOps | Linux, Git, Docker, Ansible, Terraform, AWS |
Languages โ๏ธ
Language | Level |
---|---|
Russian ๐ท๐บ | Native |
English ๐ฌ๐ง | Fluent |
French ๐ซ๐ท | Fluent |
Spanish ๐ช๐ธ | Upper Intermediate |
Italian ๐ฎ๐น | Intermediate |
Grades ๐
Grading system
The acquired knowledge and skills for different university disciplines were evaluated throughout the whole semester in the form of homework assignments, practical tests and exam. The final grade for each discipline had either one of two forms:
- Latin letters: A, B, C, D, E and F.
- Binary grade: passed or not passed.
My grades
All of my main grades for computer science disciplines are listed in the table below.
My overall GPA including all the university disciplines is 4.50 / 5.00.
Discipline | Grade |
---|---|
Algorithms and Data Structures | A |
Data Analysis and Information Retrieval | A |
Databases | A |
Database Design | A |
DevOps Technologies | Passed |
Discrete Mathematics | A |
Linear Algebra | A |
Machine Learning | C |
Mathematical Analysis | A |
Mathematical Statistics | A |
Programming Technologies | A |
Object Oriented Programming | B |
Operating Systems | C |
Software Design | C |
Software Development Tools | Passed |
Web Development | B |
Projects ๐ฅ
Machine Learning
Name | Description | Domain | Repository | Deployed app | Technologies |
---|---|---|---|---|---|
AI Chatbot | A chatbot application designed to automate hotel management using an AI agent. | Chatbot, AI agent | Frontend, Backend | - | Python, Flask, Vue.js, LangGraph, LangSmith, Whisper, MongoDB |
Food Text Generator | An application generating aggregated food description for a restaurant from several input photos. | LLM | Here | - | Python, HuggingFace, Langchain, OpenAI API |
Blackjack Strategy Optimizer | Monte Carlo algorithm implementation from scratch for finding an optimal blackjack strategy. | RL | Here | - | Python, Pandas, Plotly |
Frontend
Name | Description | Domain | Repository | Deployed app | Technologies |
---|---|---|---|---|---|
Shifumi | Implementation of the "Rock Paper Scissors" game with several modes of play. | Game | Here | Here | HTML, CSS, JavaScript, React.js, ChakraUI |
Weather app | An application displaying weather parameters for cities chosen by the user. The information is parsed from OpenWeatherMap API. | Weather | Here | Here | HTML, CSS, JavaScript, Node.js |
Achievements โญ๏ธ
Winner of the hackaton "24 heures du code"
The goal of the hackathon was to develop a chatbot to automate the reception process of a hotel. Instead of clients having to ask hotel managers for information and make requests manually, they could interact with an AI-powered chatbot that handles basic management tasks seamlessly.
The hackathon organizers provided a pre-developed REST API for the hotel system. The chatbot's objective was to interpret user requests in natural language and translate them into appropriate API calls.
In a team of three developers, I developed the backend part of a chatbot application using AI agents with the help of the LangGraph framework.
