Andrejs

Mark

A custom Telegram bot built with Python (aiogram) to automate complex data scraping, technical analysis, and AI-driven sentiment checks for Web3 markets. Instead of relying on basic price alerts, this bot orchestrates multiple APIs and local models to provide actionable data directly in Telegram. Key features of this build: API Orchestration: Integrates DexScreener for real-time liquidity pooling and GoPlus API for automated contract security audits (filtering honeypots and high-tax contracts). Automated Charting: Uses Pandas and mplfinance to calculate Market Structure Breaks (MSB) and dynamically generate chart images sent directly to the chat. LLM Integration: Connects to a AI model (DeepSeek) to scrape Twitter timelines and run sentiment analysis/fact-checking on current narratives. Database / Tracking: Implements aiosqlite to log signal setups, track simulated trailing stop-losses, calculate PnL, and maintain an ongoing win-rate database. Async Architecture

A custom Telegram bot built with Python (aiogram) that acts as an advanced algorithmic screener for Bybit futures. This project highlights my ability to combine real-time exchange APIs, machine learning, and Telegram's interface into a single automated system. The bot doesn't just forward webhooks; it independently calculates complex technical logic and manages user access. Key features built into this system: Data Engineering / APIs: Connects to Bybit to fetch live k-lines and orderbook data, calculating custom metrics like bid/ask imbalances and volume decay. Machine Learning Engine: Collects a 14-point feature vector in real-time and feeds it into a scikit-learn Random Forest model to predict setup probabilities. The system includes an auto-retraining loop based on historical trade outcomes. LLM Integration: Parses chart microstructure and sends data to a AI model to generate automated narrative analysis. Dynamic Chart Generation: Uses pandas and mplfinance to

An automated Telegram storefront designed for selling digital goods and services. Dynamic Catalog: Multi-level navigation through product categories and server selection via inline keyboards. Smart Calculator (FSM): Processes custom user inputs to calculate final prices and generate unique order IDs on the fly. User Dashboard: Tracks individual account details, purchase history, VIP status, and dynamic personal discounts. Referral System: Built-in affiliate tracking with unique invite links and automated reward calculations. Bilingual UI: Custom middleware providing real-time localization (English/Russian), saving user language preferences to the database. Checkout Architecture: Structured payment routing interface prepared for external API integrations (fiat and crypto gateways) bot code on github https://github.com/oraspid67/epg-market-bot

An asynchronous Telegram bot (Python/aiogram 3) automating bookings for service businesses. It handles complex temporal logic, real-time DB queries, and multi-step flows, replacing manual scheduling. Core Functionality: Dynamic Scheduling: Generates a rolling 7-day interactive calendar. Queries SQLite in real-time to filter available slots and prevent double-booking. State Management (FSM): Captures and validates client contact data (name, phone) to finalize reservations. Bilingual UI: Custom i18n middleware for instant EN/RU interface switching, persisting user preferences. User Dashboard: Interface for clients to track booking history and current statuses (Pending/Confirmed/Rejected). Async Architecture: Built on asyncio and aiosqlite for non-blocking performance. Integrates apscheduler for background tasks like automated pre-visit reminders. bot code on github: https://github.com/oraspid67/telegram-bots-portfolio