With the rapid rate at which climate change is progressing, California's wildfires are only getting bigger and costlier with each passing year. Fire-prone areas with high population densities, also known as the wildland-urban interface, are more vulnerable than ever: fires in these areas can start at any moment and grow massive in a matter of hours. It is vital for people living on the WUI to have real-time information indicating whether the risk of a large wildfire occurring in their area is great enough for them to evacuate. Current fire danger systems, such as the National Fire Danger Rating System, provide approximate daily fire danger ratings for an area but do not take into account constantly-changing weather conditions. To address this issue, I have created a web application where users in a WUI area can view the real-time probability of a major wildfire (1000 acres or greater) occurring in their area. I analyzed weather variables commonly associated with wildfire spread: wind speed, humidity, temperature, cloudiness, and rainfall, and found weather data from over a WUI area in the Santa Cruz Mountains where large wildfires have previously occurred. I then trained a machine-learning algorithm, logistic regression, using the dataset to determine the probability of major wildfire. Multiple runs of my model on test data-points resulted in mean precision and F1-scores of 90% and 70% respectively. Future studies will focus on improving model performance by incorporating variables like vegetation index of an area, slope grade of terrain, and types of vegetation present.