Wildfires are an escalating threat to public safety, intensified by climate change, and require effective evacuation planning to protect human lives. Developing reliable evacuation strategies necessitates a comprehensive understanding of the complex interactions between fire dynamics, human behavior, and traffic systems, along with efficient tools for large-scale scenario analysis. This research presents an integrated framework that combines traditional modeling approaches with advanced artificial intelligence models to address the challenges of wildfire evacuation planning. The framework integrates fire spread dynamics, a socio-demographic human behavior model, and traffic simulations to estimate the probability of safe evacuation for wildfire-threatened communities. Human decision-making and evacuation travel times are modeled using an agent-based stochastic approach, while traffic congestion effects are captured in detail. A Bayesian Network unifies these sub-models, enabling probabilistic inference to estimate community safety. To address the computational challenges of traditional simulations, this research proposes a Graph Neural Network (GNN)-based traffic simulation methodology. The GNN model is trained on data from a single agent-based simulation, learning the underlying traffic dynamics of the community. Once trained, it enables rapid, high-accuracy evacuation route and travel time predictions, significantly reducing computational overhead while maintaining fidelity. Additionally, this research explores the application of Large Language Models (LLMs) in disaster management. Specifically, an automated document analysis algorithm is proposed that leverages language models to extract and organize human behavior-related factors from wide range of textual data. These AI-based approaches improve human behavior modeling and contribute to a more informed evacuation planning process. The WiSE software platform integrates all these components into a unified tool for evacuation planning, tested using the 2018 California Camp Fire as a case study. By combining probabilistic modeling, AI-driven analysis, and scalable traffic simulation, this research advances wildfire evacuation planning. The WiSE platform provides decision-makers with actionable insights, helping identify bottlenecks, optimize evacuation strategies, and enhance community resilience in the face of wildfire threats.