Sensing-Seizing-Transforming: The Imperative of AI and Reinforcement Learning for Dynamic Competitiveness of Thai Brands

This study develops an AI-enhanced Sensing-Seizing-Transforming (SST) framework integrating Prophet-ARIMA hybrid forecasting and multi-agent reinforcement learning (MARL) for Thai brands' dynamic competitiveness. The methodology combines DQN-based inventory optimization (41% turnover improvement) and PPO-driven pricing strategies, validated through Monte Carlo Tree Search. Results demonstrate 8.7% profit margin growth and 19.7% market share gains during peak events, with high-end brands showing 92.3% forecast stability. The framework reduces manual intervention by 83% and improves holiday effect prediction accuracy by 38%, offering actionable insights for adaptive brand strategies in Southeast Asian markets.