Switch to English?
Yes
Переключитись на українську?
Так
Переключиться на русскую?
Да
Przełączyć się na polską?
Tak

Maksim Velichko

Offer Maksim work on your next project.

Ukraine Kyiv, Ukraine
1 day 19 hours back
Available for hire available for hire
1 proposal made
on the service 3 years
  • google sheets excel

Rating

Successful projects
No data
Average rating
No data
Rating
328
Web Programming 2
1647 place out of 6542
Data Parsing
190 place out of 784

Language proficiency level

English English: upper-intermediate

Skills and abilities

Portfolio


  • Development of a high-performance automotive marketplace parser

    Data Parsing
    Task: Create a stable and fast tool for monitoring the car market with automatic data export in Excel (CSV) format for further price analysis.

    Implemented solution and technical features:

    Asynchronous architecture: By switching to httpx.AsyncClient and asyncio, the script performs network requests without blocking the thread, ensuring high data collection speed (50+ ad pages in less than a minute).

    Data shift protection: Isolated parsing "from the object" (card-wise) has been implemented. The script dynamically checks for the presence of characteristics before writing, so the absence of price, mileage, or fuel type in individual ads does not cause errors and does not shift columns.

    Memory optimization: Data collection is optimized through fast lists of dictionaries, and the Pandas DataFrame is initialized only once at the end of the process. This eliminates system load when parsing large volumes.

    Ready business result: Data is exported to CSV with the correct utf-8-sig encoding, ensuring proper display of Cyrillic characters in Microsoft Excel without additional settings.

    Technology stack: Python, HTTPX (AsyncClient), BeautifulSoup4, Pandas, Asyncio.
  • Interactive AI Dashboard for Sales Forecasting & Metrics

    Data Processing
    Created a web application that transforms dry numbers into understandable graphs. It allows users to upload sales data and instantly receive an analysis of effectiveness (ROI, margin) and demand forecasting using ML models.

    Result: Users can make data-driven decisions without having programming skills.

    Stack: Streamlit, Scikit-learn, Plotly, Pandas.
  • Auto ML Dashboard

    AI & Machine Learning
    Interactive machine learning control panel created using Streamlit and scikit-learn. This project allows users to upload datasets, preprocess data, train multiple machine learning models, tune hyperparameters, visualize performance metrics, and export reports — all through an intuitive web interface.
  • LibraFlow

    Web Programming
    LibraFlow — is a RESTful API for managing a digital library. Users can browse, reserve, and view books, while librarians manage inventory and rentals. Features include authentication, rental limits, availability checks, user roles, ratings, and advanced filtering.

Activity

  Latest proposals 10
Data parsing through mobile API
16 USD
Parsing location map
18 USD
Parsing Google, Google Maps
22 USD
Parsing information for database creation
16 USD
Optimize parsing
16 USD
Collection (parsing) of a database with company contacts, social networks
16 USD
Parser olx/ prom.ua
22 USD
Search for a specialist in parsing (Prozorro, Clarity Project)
18 USD
Parsing
16 USD
Development of a bot
16 USD