Кирило

I have nearly a year of experience in Python development
Aizatulin Kyrylo [відкрити контакти](див. вище в блоці «контактна інформація») and data analysis. For over a year now, I have been using
Junior Data analyst [відкрити контакти](див. вище в блоці «контактна інформація»)
GitHub
Python as my primary tool, developing my own library for
rapid data analysis using Pandas and MatPlotLib, creating
I am considering remote LinkedIn intuitive and visually appealing data dashboards using
Work or working in the Certificate Power BI for schedules and Powerpoint for presentation,
city of Kyiv. and frequently working with MySQL and PostgreSQL.

Hard skils:
Programming languages: Python, SQL, С++ (bacis)
Data analysis: Pandas, matplotlib, MySQL, PostgreSQL, Power BI, Excel, Powerpoint
Machine learning: PyTorch, scikit-learn, TensorFlow / Keras, CNN, RL, SL
Web and parsing: Django, BeautifulSoup4, request, Flask
Infrastructure: Docker, Poetry, Git, Collab, Jira, Slack

Languages: Englist (upper-intermediate), Ukrainian (native)

Education and courses:
GOIT — Python Data Science and Machine Learning | aug 2025 - jan 2026 (7 months) | Python, Google Colab, SQL/MongoDB,
web scraping, EDA, statistics and a strong mathematical foundation, classical machine learning, neural networks (CNN, NLP
basics), hyperparameter tuning, time series, deployment (Dash/Django), Docker
Uzhhorod National University | 2025 - 2029 | Department of System Analysis and Optimization Theory, 2nd Year

Project experience
Analysis of warehouse data upon customer request made in just one day (Link to dashboard)
Stack (Python, Pandas, scikit-learn, PowerBI, PowerPoint)
This document contains a list of goods, a shipment log, and a list of requests for processing. Using my library, Pandas, and Power BI, I have
accomplished the following:
Initially, I reviewed the product list, identified a product with incorrectly filled-in specifications, and excluded it during subsequent
document processing.
I compiled a more comprehensive daily log (including total weight, volume, and quantity of items) by merging the original daily log
with the product list.
Identified the busiest dates based on three criteria, determined typical customer churn patterns, and ranked the top 5 customers and
top 10 products from the customer perspective.
Using the K-means algorithm, identified three primary clusters to focus on when developing a business plan.
Provided detailed business recommendations, supported by relevant charts, key metrics, and a correlation matrix. Created a visually
appealing dashboard in a PowerPoint presentation.

A search engine integrated with a PostgreSQL database (link to the GitHub repository)
Stack (Python, Pandas, Django, HTML, CSS, PostgreSQL)
The task assigned to me was to transfer the 10.000 long bank's client list, along with their profiles, into a database, and then to create a
user-friendly interface for navigating through that database. During the execution of this task, I:
Created a Python script that transfers data into a PostgreSQL database.
Connected it to a Django-based backend.
Implemented search algorithms based on specified criteria, as well as sorting functionality by range and by name.
Additionally, I built a user-friendly web interface using the same Django framework, leveraging HTML and CSS.

A chess bot being developed in three stages: the engine, SL, RL (link to the GitHub repository)
Stack (Python, Pandas, matplotlib, PyTorch(cuda), Numpy)
Starting from scratch, I developed a chess engine in Python that employs the most optimal algorithm to compute all possible moves; I then
analyzed and transformed a dataset comprising 200,000 rows into the required format, pre-trained a neural network on this dataset, and
subsequently ran a training script based on reinforcement learning (RL) technology. The development process included the following steps:
A neural network comprising two "heads" for calculating the game outcome and the optimal move;
A engine that identifies all possible moves in a given position so that the neural network can select from them;
A script that analyzes a file containing 13,000,000 lines, selects games with the required characteristics, and formats them in a format
suitable for training the neural network;
A fine-tuning script and a reinforcement learning script, both of which preserve metrics for analyzing the bot's performance;
And plots generated with matplotlib for precise data visualization.