Experience
Background
Energy Procurement Research & Backtesting Platform · proprietary platform · on-site engagement in Oxford, UK (Sep - Oct 2026)
A research and backtesting platform for power and gas procurement in European markets, built for monthly and quarterly purchases. Built in under a month. In its first exploratory backtest, the quarterly power buying rule came in 2.5% below the buyer's current practice in 3 of 3 delivery quarters (fees excluded), inside the 1 to 5% range McKinsey cites for strong energy procurement. The first 5 buying strategies are ready for validation, and the method is open-ended: new strategies and hypotheses can be added for each buying problem to find the best result. An out-of-sample period stays sealed until a rule is frozen. How it works and the evidence.
Agentic Workflow Orchestration Platform (AWOP) · proprietary platform
Berlin, remote. A platform for AI coding agents. It develops the trading decision engine and the execution copilot today, and built the energy research platform. Several AI agents work in parallel under one shared engineering standard, with a test-regression gate on every merge and independent review on every autopilot merge. How it works and the evidence.
Trading Execution Copilot · voice execution on MT5 · proprietary product
A voice execution copilot for discretionary trading on MT5, built through the same agentic platform. Every order is validated, risk-sized and read back; nothing is sent without explicit confirmation, and every execution is reconciled with the broker. How it works and the evidence.
Independent quantitative research and trading-systems development
Berlin. Intensive self-directed market study and applied research, through four successive trading systems. Self-funded throughout.
Trading-strategy training
Seven successive private mentorship programmes in discretionary trading. 564 archived sessions, and 15,000+ of those hours in chart study. Those rules were then formalised into deterministic, testable logic, and the ones that did not survive measurement were dropped.
Quant research and algorithm optimisation · Rabot Crypto
Data-driven research for crypto markets. Python dashboards with pandas, NumPy, Dash and Plotly for market behaviour, volatility regimes, volume and open interest. Translated research ideas into testable workflows with developers and traders.
Data Science and Machine Learning bootcamp · SPICED Academy
Berlin. 480+ curriculum hours: Python, pandas, NumPy, scikit-learn, SQL and NoSQL fundamentals, dashboards, deployment concepts.
Music producer, educator and founder
Berlin and international. Built a creative practice and an online education business: production, teaching, mentoring and independent ownership. It is where the discipline of structure, timing and iteration comes from.
The three systems before this platform: a configurable multi-timeframe BTC research system in June 2025, a Python and Streamlit application for rule-based signal research in September, and a React and TypeScript platform with engine comparison and golden-case validation in November.
Core expertise
Looking for someone who has already built the whole thing?
Every figure on the site is traceable: the run manifest, the git history, the test directory. Ask me for any of them.
