This project is a focused experiment in running a small, aggressive AI Buildout Moat Portfolio rather than a broad AI basket. The goal is to find companies with real infrastructure, software, power, security, automation, defense, or digital-system exposure that can benefit as AI adoption scales.

New ideas are judged on direct AI buildout exposure, a specific moat, upside asymmetry, financial quality, growth, valuation, catalysts, technical setup, risk, and portfolio fit. The account favors names where the market may still be underestimating the value of the company’s AI position.

Position sizing is judgment-based, not formula-based. Small exploratory positions are acceptable for developing theses, while larger weights require clearer evidence, better risk-reward, and stronger conviction. Concentration is allowed when the upside case is strong, but hidden theme overlap must be flagged.

Risk management is driven by thesis quality, concentration awareness, and downside reviews. A roughly 10% drop from cost triggers review, 15% triggers a serious thesis check, and 20% forces a hold, trim, or sell decision with explicit reasoning.

The project can recommend buys, adds, trims, and rotations, but it cannot place, cancel, or modify a live trade without explicit user confirmation for that specific action. The site tracks the experiment, the watchlist, the decision process, and the current state of the account rather than acting as an execution engine.