Olga S.
Offer Olga work on your next project.
Private message
Offer work
Brovary,
Ukraine
8 months 17 days back
available for hire
on the service 8 months 18 days
Rating
Successful projects
Average rating
Skills and abilities
Portfolio
-
338 USD AI upscaling of images and videos
AI & Machine Learning
Key technologies and architecture:
- AI models: Integrated Real-ESRGAN (x2/x4/x8, normal/anime modes) for upscaling and GFPGAN (v1.4) for face restoration.
- Model management: Implemented Singleton ModelManager with LRU cache (based on OrderedDict) for efficient model loading and switching.
… - GPU optimization: Includes GPUCapabilities class for automatic detection of VRAM and compute capability, allowing dynamic adjustment of tile_size, half-precision (FP16), and enabling TF32 for Ampere/Ada GPUs.
- Multithreading: Image processing moved to QThread (UpscaleWorker) to prevent GUI blocking.
- Video processing: Implemented pipeline (reader/processor/writer) on threading.Thread and queue.Queue for parallel frame processing. Uses subprocess to call ffmpeg (demultiplexing, saving audio tracks, assembly).
- Filters: Modular FilterPipeline for applying a chain of filters (CLAHE, Bilateral denoise, Dehaze, Canny edge sharpening, etc.).
- Memory optimization: MemoryOptimizedProcessor for processing ultra-large images (tiling).
Functionality:
- Batch image processing (including Drag-n-Drop).
- Live preview with SplitView widget (comparison "before/after").
- Preset management (via QSettings).
- Localization (JSON, Translator) and theme switching (QSS).
- Real-time monitoring of VRAM/RAM (psutil, pyqtgraph).
-
135 USD TGStat parser with graphical interface
Data Parsing
Desktop application in Python for collecting data on Telegram channels and chats from the TGStat website. The application features a full graphical interface implemented in PyQt6 and uses the DrissionPage library to control the Chromium browser and parse web pages.
Key features:
… - Graphical User Interface (GUI): An intuitive interface in PyQt6 allows easy configuration of collection parameters, starting/stopping the process, and monitoring execution logs in real-time.
- Browser Management: DrissionPage is used for browser automation, supporting both regular and headless modes.
- Cloudflare Bypass: A class is implemented for automatic passage of basic Cloudflare checks ("Just a moment...").
- Authentication Support: The application checks for an active session on TGStat and, if necessary, waits for manual user login, saving the profile for future runs.
- Two parsing modes:
- By Categories: Flexible selection of countries and categories for bulk data collection.
- By Links: Collecting information from a provided list of direct URLs.
- Filtering: Ability to select type (channels/chats) and set a minimum threshold for the number of subscribers.
- Multithreading: Browser connection and parsing tasks are executed in separate threads (QThread), preventing the interface from freezing.
- Data Export: Collected data (name, subscribers, link, category, etc.) is automatically saved to an .xlsx file using pandas.
Stack: Python, PyQt6, DrissionPage, pandas.
-
226 USD Multifunctional Telegram bot for marketing
Bot Development
Multifunctional Telegram bot for marketing
A multifunctional Telegram bot in Python designed for marketing automation. The architecture includes two components: a network of user-bots on Telethon and a management admin panel on Aiogram.
… Key functionality:
- Multi-account core: Management of a pool of Telethon accounts with proxy support (including rotation, bulk addition, and auto-assignment).
- AI generation: Integration with OpenAI API (with a key rotation mechanism, limit consideration, and fault-tolerant fallback to G4F) for generating comments and conducting dialogues.
- Automation:
- Responding to keywords (global and by categories).
- Auto-commenting on new posts in target channels.
- Conducting multi-step dialogues with users.
- Sending comments on a schedule (APScheduler).
- Admin panel: Full UI for managing accounts (checking, removing banned ones, bulk updating profiles), chats (bulk adding/removing), AI settings, analytics (SQLite), and background tasks.
-
45 USD Universal Media Downloader
Python
Universal Media Downloader is a cross-platform desktop application for downloading media (video and audio), written in Python using the PyQt6 framework.
The application is built on a multithreaded architecture (QThreadPool, QRunnable) to ensure a fully responsive UI during operations. It uses yt-dlp as the backend to support hundreds of sites (YouTube, VK, TikTok, RuTube, etc.) and FFmpeg for post-processing (e.g., format merging or audio extraction).
… Key features:
- Parallel downloading of multiple files with queue management (start, stop, retry task).
- Flexible input: direct URL paste, batch paste from clipboard, drag-and-drop links and txt files.
- Advanced settings:
- Quality selection (including "audio only") individually for each service.
- Cookie support (from file or through browser auto-detection) for accessing private content.
- Automatic subtitle downloading.
- Modern UI:
- Custom widgets (QListWidget) for displaying list items (preview, progress, status).
- Full theme customization (Light/Dark) via QSS.
- Internationalization (i18n) with on-the-fly language switching (Ru/En/Uk).
- Reliability: logging, ffmpeg presence check at startup, cleanup of temporary files on cancellation.
Activity
| Latest proposals 4 | Budget | Added | Deadlines | Proposal | |
|---|---|---|---|---|---|
|
Telegram bot for courses/subscription
79 USD
|
|||||
|
Telegram bot for access to the online course
90 USD
|
|||||
|
Download the video course from my account
68 USD
|
|||||
|
Telegram post grabber
16 USD
|
