I build algorithmic trading systems and quantitative research infrastructure, with 12+ years of hands-on experience in trading software (including deep MT4/MT5/MQL5 domain work).
Focus: turning trading ideas into testable, reproducible, production-oriented systems.
About the name. NeoZorK is my GitHub handle for software engineering work. It is not connected to any video blogger, streamer, gaming channel or game account that uses the same or a similar nickname. This profile contains only professional work: trading systems, research tooling and open-source software.
- Algorithmic trading and market research
- Python / C++ trading infrastructure
- Quantitative validation and backtesting
- AI-assisted R&D (as a development multiplier, not a job title)
- Proprietary market analytics
| Layer | What |
|---|---|
| Core | Algorithmic trading · trading systems · Quant R&D |
| Strong | Trading architecture · validation · MQL5/C++ domain depth |
| Applied | Python · FastAPI · PostgreSQL · Docker |
| Emerging | ML / LLM tooling · local AI workflows · MLX |
- Monte-Neo — an independent verifier for trading strategies written by AI agents and humans. It looks for look-ahead bias, hidden trading costs and overfitting before a backtest reaches real money, and issues reproducible, signable certificates. Adapters for common backtesting frameworks; MCP server for agents.
- ClaimBound Evidence — preregistered evidence discipline
Green /
PASSED_UNDER_PROTOCOL≠ independently reproduced. Most public cards remain single-operator until a separate rerun.
Kept public for reference, read-only, and described as they are. None of them is a working trading product.
| Repository | What it was | What is actually there |
|---|---|---|
| trading-data-replay-engine | A two-hour engineering exercise: replay of historical and live quotes | Latency-aware replay order (timestamp + latency), Redis queue, mid-price processor. The arrival-time idea continues in Monte-Neo's verification checks. |
| DEXArb | EVM DEX pool scanner (C++) | Multi-threaded factory and pool scan over public JSON-RPC. No arbitrage detection, execution or wallet logic. |
| NeoZorK3 | Solana arbitrage-bot skeleton (C++) | CLI, configuration, RPC endpoint discovery. The arbitrage engine and transaction signing are stubs. |
| neo-slack-bot-production | Personal Slack Socket Mode client (C++) | Socket Mode connection with native macOS notifications; v0.0.7, not battle-tested. |
I use modern AI coding/reasoning systems (Claude, GPT, Gemini, Grok, Cursor, local LLMs) to accelerate research iteration, refactoring, tests, experiment automation, and trading-system prototyping — not as “generic AI development.”
Open to remote Quant Developer / Algorithmic Trading / Trading Systems R&D roles.
GitHub: @NeoZorK