Martin Poulsen@mpoubot
1
Events attended
1
Submissions made
Denmark
10+ years of experience
About me
I'm an experienced Project Manager and technology consultant with 15+ years of experience delivering complex IT and business transformation projects. Over the last few years my focus has shifted increasingly toward AI, automation and building systems that can turn data into decisions and actions. I'm particularly fascinated by the point where AI moves beyond answering questions and starts operating as an autonomous agent — researching, evaluating alternatives, making decisions, executing actions and learning from the results. I'm currently applying that thinking to algorithmic trading. I'm building a research and trading platform that combines market data, technical analysis, asset selection, liquidity, execution, AI analysis and risk management. Rather than jumping directly from an idea to live trading, I'm building a structured lifecycle from research and backtesting through out-of-sample validation, paper trading, live validation and ongoing monitoring. For the Alpaca hackathon, I'm taking this further into autonomous options trading. My goal is to build something that is not just an AI trading demo, but a complete decision-making system that can explain why it sees an opportunity, determine whether the trade is executable, manage risk and evaluate its own performance. I enjoy experimenting, challenging assumptions and turning ideas into working systems. I'm here to build, test, learn — and hopefully create something genuinely useful along the way.
🤝 Top Collaborators
Ilze Rosicka
🤓 Latest Submissions
DELTAX V2 — Autonomous AI Trading Agent
DELTAX V2 is an autonomous AI trading system built for the Alpaca AI Trading Agents Hackathon. It combines technical market analysis, real-world news intelligence, deterministic risk controls, automated broker execution, and end-to-end auditability into a single trading architecture. The system operates through three distinct trading strategies: Core, Active, and Intraday. Technical analysis identifies potential opportunities first, while AI evaluates whether current news, market conditions, confidence, and directional context support or invalidate the trade thesis. Before execution, every candidate must pass deterministic risk gates covering AI confidence, conflicting news, market and sector conditions, technical confirmation, timing rules, existing positions, and portfolio exposure. DELTAX limits planned loss to 1% of a $100,000 paper portfolio per trade, disables new entries at a 3% daily drawdown, and activates a kill switch at 5%. For options, DELTAX uses defined-risk credit spreads with systematic contract-selection rules. Orders are executed through Alpaca Paper Trading, reconciled against broker state, and continuously monitored. Every decision is recorded in PostgreSQL, creating a traceable path from technical signal → AI confirmation → risk gates → trade intent → Alpaca order → position → exit. A read-only dashboard exposes decisions, P&L, system state, and execution history for transparent review. DELTAX V2 demonstrates how AI can go beyond generating trading signals by acting as a disciplined validation, risk, execution, and auditing layer around an autonomous trading system.
4 Sep 2026
