Eugen Soloviov
Trading-systems engineer
Ingénieur en systèmes de trading construisant des bots depuis 2017 : arbitrage inter-échanges (connecté jusqu'à 30 plateformes), arbitrage de paires basé sur la cointégration entre spot et futures, scalping, stratégies basées sur l'actualité et le sentiment, algorithmes de tendance, et algorithmes de gestion et d'équilibrage de portefeuille. Construit également l'exécution d'ordres en sous-milliseconde, des entrepôts de big data, des moteurs de backtesting, des agents IA et des interfaces de trading (y compris profitmaker.cc open-source). Stack : JS/TS, Python, Rust/Zig/Go, DevOps, backend, frontend, architecture.
Articles
Backtest-live parity: why your bot trades differently from the backtest
Complete taxonomy of divergences between backtesting and live trading: from slippage and partial fills to codebase desynchronization. Architectural patterns for achieving parity, Python examples of a shared core module, and a production monitoring checklist.
Monte Carlo Bootstrap: How to Get Confidence Intervals for a Backtest in 10 Lines of Code
Why a single-point estimate from a backtest is a dangerous illusion. How Monte Carlo bootstrap in 2 seconds of computation gives you a 95% confidence interval for PnL and MaxDD, and why this is a mandatory step before launching a strategy in production.
Funding Rate Arbitrage Across Exchanges: How to Profit from Rate Differences
How funding rate arbitrage works across crypto exchanges, why rates differ on Binance, Bybit, OKX and dYdX, and how to build a monitoring and execution system to extract profit from these discrepancies.
QuestDB for Algorithmic Trading: SQL Extensions That Change the Game
Deep dive into QuestDB's time-series SQL extensions: SAMPLE BY, ASOF JOIN, HORIZON JOIN, WINDOW JOIN, LATEST ON, and real-world trading query patterns.
QuestDB for Algorithmic Trading: From Order Books to Production Architecture
Materialized views, 2D array order book analytics, and reference architecture for a QuestDB-powered algorithmic trading platform.
QuestDB for Algorithmic Trading: Architecture That Speaks the Language of Markets
Deep dive into QuestDB's three-tier storage architecture — WAL, columnar storage, and Parquet on object storage — and schema design principles for algorithmic trading systems.
Data Communication in Algo Trading Systems: A Technology Overview
We analyze communication technologies at all levels of an algorithmic trading platform: from exchange connectivity protocols (REST, WebSocket, FIX) to internal IPC, message brokers, and data stores.
Loss-Profit Asymmetry: The Math That Kills Your Deposit
Why losing 50% requires 100% growth to recover, how volatility drag destroys capital even in sideways markets, and which formulas every algo trader must know for building risk management.
Complex Arbitrage Execution in Rust: From Nanoseconds to Atomic Multi-Legs
How to squeeze maximum performance out of Rust for multi-leg arbitrage execution: io_uring, lock-free order books, LMAX Disruptor, SIMD, type-state machines, and arena allocators.
GNN, Transformers, and RL for Arbitrage: When Neural Networks Learn to Trade
How graph neural networks find arbitrage chains in 78 ms, why RL agents show 142% annual returns against 12% for rule-based bots, and how to build an integrated system in Rust.
Matrices, Tensors, and Tropical Algebra: Linear Algebra for Arbitrage Detection
How the matrix of exchange rates, eigenvalues, tropical algebra, and tensor decompositions turn cryptocurrency market chaos into clear arbitrage signals.
Vine Copulas for Arbitrage: Modeling High-Dimensional Dependencies
How to use Vine Copulas to identify hidden dependencies between dozens of crypto assets and build robust, high-dimensional statistical arbitrage strategies.