Dmitro Vilihotsky
Offer Dmitro work on your next project.
Rating
Language proficiency level
Skills and abilities
Programming
Services
Mobile development
Portfolio
-
223 USD Crypto Trading Bot - pump screener and Open Interest in real time
Bot DevelopmentTelegram bot that monitors the cryptocurrency futures market 24/7 and instantly catches sharp movements - earlier than most traders will see them.
Problem:
A trader cannot physically monitor hundreds of coins simultaneously. A pump on a small coin lasts minutes: by the time you scroll through the terminal, the movement has already occurred.
…
What has been done:
An asynchronous bot in Python, connected to two exchanges - Binance Futures and Bybit - via official APIs and WebSocket streams. The bot processes the price stream in real-time for hundreds of USDT pairs (in the production version - 577) and sends a signal to Telegram seconds after the event.
Two independent scanners
— Pump scanner: detects a sharp price change over a specified interval (threshold and timeframe are adjustable)
— Open Interest scanner: tracks the influx of money into positions - an increase in OI often precedes price movement.
Signals include entry point, stop-loss, take-profit, risk/reward, and confidence level. There are quality filters and anti-spam measures: no more than 3 signals per coin per day to avoid flooding the chat with noise.
Technical details:
Fully asynchronous architecture (asyncio + WebSockets) — one process handles hundreds of data streams without delays. The bot menu features inline buttons: personal scanner settings, statistics, test signals. Configuration via environment variables, ready for deployment on VPS.
Stack: Python (asyncio), python-telegram-bot, Binance API, Bybit API, WebSockets, pandas.
Important: the bot provides analytical signals, not financial advice.
-
446 USD ContentFlow - Telegram Mini App for auto-posting videos on social media
Bot DevelopmentContent automation platform: upload a video once, choose the time - and it publishes itself. Everything is managed directly from Telegram.
Problem:
The content creator is tired of manually posting the same short videos daily on YouTube, TikTok, and Instagram. They needed one tool instead of three tabs and an alarm.
…
What has been done:
A full-fledged Telegram Mini App built from scratch. Frontend on React - a clean mobile-first interface right inside Telegram. Backend on FastAPI on a cloud VPS: database, video processing, publishing. A background scheduler posts each video exactly at the time chosen by the user. The app pulls videos from a link, queues them, has a publication calendar, custom captions, and a posting schedule - all in one place.
The most challenging technical challenges
— Real OAuth authorization for YouTube: secure connection to the user's channel, automatic token refresh (the connection never drops), uploading via the official API.
— Complete server setup independently: Linux VPS, HTTPS with a real domain, reverse proxy, database - stable operation 24/7.
— Time zones: the server lives in UTC, but a post scheduled for "18:00" must go out at 18:00 according to the user's clock. I rebuilt the scheduler so that the time always matches the local time.
Result:
Connected the channel once → dropped the link to the video → chose the time → went about their business. Everything else is automatic.
Stack: React, FastAPI (Python), Telegram Mini Apps, YouTube Data API, OAuth 2.0, PostgreSQL, Linux VPS, Nginx.
-
892 USD NetGuardian - an AI network monitoring system with automatic removal
PythonIntelligent monitoring and auto-recovery network system that I designed and built entirely on my own - from software to hardware.
Graduation project in the specialty "Computer Engineering".
Problem:
… The internet disappears during an important call or game - and no one understands why: router, provider, DNS, or the computer itself? NetGuardian works as a personal network engineer: it accurately identifies the source of the problem and resolves it automatically.
Architecture - three independent components
— Desktop agent (Python, CustomTkinter): 9 modules - real-time diagnostics, Wi-Fi analysis, DNS benchmark, local network security
— Hardware agent on Raspberry Pi: autonomous 24/7 monitoring via MQTT, independent of the computer
— Telegram bot: 14 remote control commands and automatic incident notifications
The main feature - hybrid AI engine with three levels
Online it operates through Google Gemini, without the internet it switches to a local database with 200+ scenarios, and for a "second opinion" it queries the Pi agent. The system always provides an answer - even during a complete network failure.
The most challenging task
To maintain functionality without the internet. Solved with offline architecture featuring redundancy and 35+ scenarios for automatic fixes: DNS reset, IP update, MTU repair, etc.
Result:
Ready standalone Windows application (24.5 MB), tested in real home networks.
Stack: Python, Google Gemini API, MQTT, Raspberry Pi, SQLite, Telegram Bot API, PyInstaller.
-
223 USD sKladMan - adaptive online clothing store
HTML & CSSProblem:
A clothing store was selling entirely through Instagram: orders via direct messages, manual inventory tracking in Google Sheets. Orders were getting lost, sales were being missed. The task was to create a fast web store based on their existing product data.
What was done
A high-performance e-commerce application built from scratch. Mobile-first interface: sticky bottom navigation, sliding filters, slide-out cart, language switch UA/EN. The frontend dynamically processes products, sizes, and prices based on inventory data.
…
Most interesting technical solutions
— Three-step modal order funnel: contacts → smart delivery with interactive auto-completion of Nova Poshta and Ukrposhta branches → payment selection. The checkout feels natural for social media shoppers.
— Fully asynchronous filters and sorting: by sizes, categories, promotional items - instantaneously, without page reload, fast even on budget smartphones.
Result
The store moved away from manual sales via direct messages. Customers see real-time inventory and can place orders in less than 45 seconds from any phone.
Stack: React, JavaScript, LocalStorage, responsive design.
Activity
| Latest proposals 7 | Budget | Added | Deadlines | Proposal | |
|---|---|---|---|---|---|
|
Python programmer
112 USD
|
|||||
|
A Telegram bot is needed. The programming language is Python, and the bot must have an admin panel for processing requests.
45 USD
|
|||||
|
Development of a Telegram bot
223 USD
|
|||||
|
Online clothing store
156 USD
|
|||||
|
A bot needs to be created in Telegram for subscription payment.
45 USD
|
|||||
|
Website design update
22 USD
|
|||||
|
Creation of a business card website for a consulting company
100 USD
|