Q-GAP (Quantum-Inspired González Adaptive Portfolio) is a physics-inspired framework for dynamic portfolio optimization with adaptive machine learning calibration. The framework models financial markets as evolving complex systems and derives portfolio allocations through a proprietary dynamic optimization process inspired by theoretical physics. An adaptive machine learning layer characterizes market conditions and supports the calibration of the model’s internal configuration. This public summary presents the scientific concept, general architecture and empirical validation of Q-GAP over a diversified universe of seven UCITS ETFs. The validation includes dynamic backtesting with look-ahead-free portfolio-weight computation, walk-forward analysis, market stress tests, sub-period analysis, block-bootstrap Monte Carlo simulations and a fully walk-forward protocol in which the machine learning layer is retrained monthly using only past data. For the January 2017–June 2026 dynamic backtest, Q-GAP achieved an annualized return of 17.40%, a Sharpe ratio of 1.067, a Sortino ratio of 1.322, a Calmar ratio of 0.896 and a maximum drawdown of −19.42%. Under the stricter fully walk-forward protocol, it achieved an annualized return of 17.06% and a Sharpe ratio of 1.037. The exact mathematical formulation, governing equations, internal parameters, calibration procedures, source code and implementation details required to reconstruct the Q-GAP engine are intentionally withheld and protected as trade secret material under Directive (EU) 2016/943. This record contains two files:1. English Public Summary.2. Spanish Public Summary / Resumen Público en español. This document has not been peer reviewed. The reported results are historical and do not constitute investment advice, a financial recommendation or a guarantee of future performance. Institutional licensing and confidential technical evaluation:qgap@mfgonzalez.commfgonzalez.com