6+ years transforming complex data challenges into elegant, scalable solutions. From real-time IoT platforms to intelligent AI assistants.
Creating focused products that augment human intelligence and automate tedious work, built with FastAPI, React, and modern AI APIs.
Your personal prompt engineer. Transforms vague ideas into precision-engineered prompts that unlock the full potential of any AI model. Built with FastAPI backend, OpenRouter API integration, and token counting optimization.
Your intelligent filter for internet noise. Chrome extension that extracts Reddit thread content and generates AI-powered summaries instantly. Handles CORS, manifest v3 configuration, and real-time content extraction.
Beat procrastination with your personal AI coach. Never feel stuck again! Chat with your board to break down overwhelming projects, get motivation when you need it, and stay on track. Your AI buddy helps you start, guides you through, and celebrates your wins.
Your Socratic co-author. Overcome writer's block by turning your raw thoughts into a publication-ready article through a guided AI interview. Elicito AI doesn't write for you; it pulls the expertise out of you, ensuring every piece is 100% in your authentic voice.
Three abilities that turn "it can't be done" into production code
Give me confusing PDFs with no text layer, scattered data across 30 websites, or millions of unstructured records, I'll find the hidden patterns and build a system that extracts exactly what you need. Chaos becomes structure.
Node.js project with zero experience? Mastered it in weeks. New database type? I'm productive day one. Unfamiliar API? Integrated by afternoon. I don't memorize syntax, I understand underlying patterns that work across all technologies.
I don't stop at surface-level implementation. I dive into PostgreSQL index internals, PDF file structure, browser detection mechanisms. This deep understanding lets me find solutions in places others don't know exist.
From scraping and parsing to enterprise AI platforms, solving fundamental business problems with elegant code.
Designed a high-throughput Azure infrastructure using partitioned Service Bus queues and TTL indexing in Cosmos DB.
Enabled AI insights from previously inaccessible image-based content through a novel in-document translation and analysis method.
Architected a large-scale RAG platform on Azure for intelligent document analysis. Engineered complex data ingestion pipelines from Google Drive, Outlook, and IMAP, and developed a unique system for translating and analyzing both text and images within PDF files using AI Vision.
Reduced test execution from 2 minutes to 5 seconds, optimized Docker builds, automated Chromium VM lifecycle to reduce costs
Built a platform integrating Google Calendar, Gmail, and Slack with ChatGPT agents. Created Chrome extension for page-context interaction, implemented OAuth2 flows, and managed full GCP deployment with CI/CD pipelines.
Delivered a major feature on an aggressive timeline, quickly becoming the team's subject-matter expert in Node.js.
Successfully implemented domain-level OAuth, allowing entire organizations to onboard seamlessly.
Quickly mastered unfamiliar Node.js stack to build real-time calendar synchronization service. Integrated Google and Outlook calendars with Auth0, implemented WebSockets for live updates, and solved complex recurring event patterns.
Managed terabytes of time-series data with TimescaleDB and created a dedicated microservice for stream lifecycle management.
Achieved a 40% reduction in infrastructure costs through architectural optimizations.
Designed and deployed entire backend infrastructure to process real-time data from thousands of IoT sensors. Integrated high-performance CCTV stream decoding accessible to multiple authenticated users simultaneously across multiple cities.
Enabled non-technical clients to monitor and control sophisticated automation systems through an intuitive interface.
Implemented dual-database architecture maintaining data isolation while providing unified access layer.
Built REST API dashboard in Django REST Framework for real-time bot monitoring, settings management, and contact data export. Configured database router to separate dashboard and bot databases. Integrated noVNC for real-time system screen broadcasting to client dashboard.
Achieved undetectable automation by emulating human behavior at operating system level rather than browser automation APIs.
Deployed on DigitalOcean VM with graphical desktop environment enabling real browser operation.
Engineered sophisticated bot for automated LinkedIn prospecting and outreach. Local FastAPI server coordinated with browser extension to emulate human behavior at OS level using mouse movements and keystrokes, preventing detection. System searched for relevant companies and decision-makers, performed initial communication, and collected data while mimicking natural user patterns.
Extracted structured data from highly unstructured sources including scanned documents and image-based PDFs.
Built robust multi-source data collection system handling diverse formats and access restrictions across European real estate markets.
Scraped and parsed building contract data from 60+ European websites including poorly structured PDFs. Implemented OCR for text recognition in images and PDFs, coordinate extraction, proxy rotation, request spoofing, and browser behavior emulation. Integrated Google API for address geocoding and organization enrichment.
A collection of smaller projects and freelance work demonstrating a wide range of skills.
LinkedIn automation bot emulating human behavior at the OS level with realistic mouse trajectories and keyboard timings.
Scraped and parsed construction contracts from 60+ sources including complex multi-column PDFs with OCR and coordinate-based extraction.
Built a search engine for millions of patents, using custom PostgreSQL RUM indexes to deliver full-text search results in milliseconds.
System to parse webpage text and generate AI-based ad copy using Markov chains and async paraphrasing with NLP services.
Created an image matching service to find visually similar products in an e-commerce database using computer vision.
Automation tools and chatbots for Telegram, Facebook, and Instagram with IBM Watson NLP integration for content analysis.
Developed a Viber bot to track employee work hours by generating and scanning encrypted, time-sensitive QR codes.
Developed scripts to parse video data, comments, and channel information, implementing proxy and user-agent rotation to bypass blocks.
Expert-level proficiency in the Python ecosystem with strong capabilities across the full backend stack, cloud infrastructure, and specialized data domains.
Master of Computer Science from National Technical University "Kharkiv Polytechnic Institute"
I chose programming because it felt like magic, the ability to create something from nothing and shape the digital world. Over 6+ years, I've evolved from scraping and parsing specialist to full-stack backend architect building enterprise AI platforms.
What drives me: I'm passionate about making things work optimally. Whether it's reducing query times from 30 seconds to milliseconds, or building Chrome extensions that seamlessly integrate with AI models, I love transforming complex challenges into elegant solutions.
My approach: I prefer working in small teams where I can see direct impact. My tenacity and ability to analyze large amounts of information enable me to solve even the most complex and ambiguous problems. I'm known for quickly grasping new technologies, once a CTO trusted me with a Node.js project despite my lack of experience, and I mastered it fast enough to lead a major feature.