The growing adoption of solar energy presents both an opportunity and a challenge. While clean and sustainable, solar power generation is inherently variable due to changing weather conditions, making reliable forecasting critical for grid integration, infrastructure planning, optimal solar panel placement, and overall energy stability. Solaris AI addresses these challenges by integrating Generative AI with traditional Deep Learning models, creating an intelligent and explainable expert system for solar energy analysis. The system initially featured an Artificial Neural Network (ANN) model trained on key meteorological parameters such as cloud cover, humidity, temperature, and wind speed to predict solar power output. It was later enhanced with a Long Short-Term Memory (LSTM) model, which is better suited for handling time-series patterns in solar power output. To make these insights accessible to a wider audience, Solaris AI incorporates Retrieval-Augmented Generation (RAG) using OpenAI’s language models and LangChain. The model, along with the dataset, is integrated into the system to generate explainable and accurate forecasts through natural language interaction. This enables stakeholders, including government planners, policymakers, businesses, grid operators, and citizens, to interact with the system conversationally, retrieve relevant data, and receive model predictions. A clean web interface with cloud deployment ensures scalability and usability. By bridging the gap between technical forecasts and practical decision-making, Solaris AI supports the broader adoption and integration of solar energy worldwide.