Accurate day-ahead electricity price forecasting is essential for market participants to optimize bidding strategies and manage financial risks in increasingly volatile energy markets. This paper presents a comprehensive web-based system designed for day-ahead electricity price forecasting based on gradient boosting models, specifically applied to the Croatian market. The proposed approach integrates a multi-layered architecture that includes data ingestion from ENTSO-E and Open-Meteo, a robust backend for model selection, training and management, with an interactive frontend presentation layer. The performance of two state-of-the-art gradient boosting algorithms, XGBoost and LightGBM was evaluated using hourly DA market price data from 2022 to 2024 from Croatian electricity market. The results demonstrate that both models achieve similar accuracy for the year 2024. Monthly analysis reveals significant performance variations, with higher errors during periods of extreme volatility and price spikes, such as summer and late winter. The study highlights the importance of meteorological features and market fundamentals in capturing price dynamics.