Maksym Ustymenko@makss197480
3
Events attended
3
Submissions made
Ukraine
1 year of experience
About me
Vibe-coding using AI tools has recently become an exciting, important, and priority activity for me! I started with the development of robots for trading on cryptocurrency exchanges and quickly came to understand the need to create a system for designing and constructing crypto trading bots. Thus, the Commander project was born — a system designed to manage trading processes using crypto bots. It was created entirely using AI-driven code generation methods. Commander is intended to become part of a complete ecosystem for designing, testing, and managing cryptocurrency trading bots. I hope that it will provide not only personal practical value but also be useful to many users interested in the cryptocurrency field.My focus is on making complex algorithmic trading accessible through intuitive AI-orchestrated workflows.
🤓 Latest Submissions
MAITS-REILLY: Multi-Agent Intel Tracking System
MAITS-REILLY (Multi-Agent Intelligence Tracking System) is a next-generation autonomous intelligence engine designed to transform exhausting manual web research into a rapid, automated pipeline. Today, the internet is full of data, but extracting it at scale is incredibly difficult. When you try to gather fragmented information across multiple pages—like manually clicking through dozens of hackathon team profiles just to find their GitHub repositories—websites quickly detect unusual activity. They block access, throw endless CAPTCHAs, or deploy anti-bot protections. Manual data gathering wastes hours, while scripts get banned. How does MAITS-REILLY solve this? The system combines artificial intelligence with advanced proxy infrastructure, operating in three key stages: Autonomous Navigation (Maxwell Daemon): Our AI agents independently map out the search strategy. They can navigate deep into websites, analyze page structures, and extract hidden data that standard search engines or basic scrapers miss. Absolute Stealth: By integrating with enterprise proxy networks (like Web Unlocker), the system dynamically routes traffic and mimics real users from different countries. To website security systems, MAITS-REILLY looks like thousands of regular human visitors, ensuring uninterrupted data collection. The Trust Gate: The biggest risk of modern AI is "hallucinations" (inventing facts). To eliminate this, we developed a dedicated verification module. It cross-references the gathered raw data against primary sources, guaranteeing that only 100% factual and grounded information makes it into the final analytical report. Ultimately, the user gets a powerful tool with a sleek, secure Command Line Interface (CLI). You simply enter a target query, and MAITS-REILLY does the heavy lifting: it bypasses defenses, deep-crawls sources, analyzes the data, and delivers structured insights. It is the perfect, time-saving solution for OSINT analysts, researchers, and data engineers.
31 May 2026
VORTEX | Multi-Agent Cross-Asset Intelligence)
VORTEX is an autonomous, multi-agent intelligence ecosystem engineered to master the high-volatility interface of crypto-assets and tokenized equities (xStocks) on the Kraken exchange. Moving beyond rigid, single-strategy algorithms, VORTEX employs a decentralized 'Council of Advisors' architecture. This council consists of seven specialized, Explainable AI (XAI) agents, each analyzing the market through distinct mathematical and fundamental lenses (Macro, Spectral, Fractal, Sentiment, etc.). A central 'Director' agent aggregates these distinct signals using a performance-based Reinforcement Learning (RL) feedback loop, dynamically updating agent weights ($w = accuracy^\beta$) in real-time. Critically, VORTEX prioritizes strict capital preservation through integrated, volatility-aware risk management, featuring ATR-adjusted position sizing, confidence-scaled execution, and adaptive circuit breakers. VORTEX is not just a tool; it is a scalable, enterprise-ready solution for the future of hybrid digital finance.
19 May 2026
ARCEON: Autonomous Agentic Economy on Arc
ARCEON is a sophisticated experiment in decentralized automation, designed to showcase the future of the Agentic Economy. At its core, the project moves beyond traditional automation by creating a "digital ecosystem" where AI agents act as independent economic participants rather than mere scripts. The system orchestrates a swarm of five autonomous agents, each with a distinct role: The Architect (Manager), and specialized dealers for Energy, Compute, Data, and Storage. These agents are programmed with a "survival instinct" tied to their financial balance. To continue their operations, they must successfully trade resources with each other, creating a true machine-to-machine (M2M) marketplace. Technologically, ARCEON leverages the Arc Network for near-instant transaction finality and Circle Nanopayments for secure, dollar-denominated settlement in USDC. This combination allows for high-frequency microtransactions that would be impossible on traditional rails due to high gas fees or latency. The project demonstrates a scalable model for future smart cities and decentralized clouds, where infrastructure components—like solar grids, storage clusters, and compute nodes—can autonomously negotiate, purchase, and settle their resource needs without human intervention. ARCEON isn't just a trading bot; it's a blueprint for a self-sustaining digital society.
26 Apr 2026
