In 2020 alone, COVID-19 caused over 30 million cases and 550 thousand deaths solely in the U.S. One of the most powerful tools that the 21 century uses is AI technology and regression modeling for prediction networks yet few neural network models are used for the worldwide pandemic. It was hypothesized that a recurrent neural network (RNN) combined with an SVM regression model can be used to analyze existing COVID cases, deaths, and hospitalizations as a base to predict future outcomes. The researcher also hypothesized a decline in COVID cases, hospitalizations, and deaths. In order to build the recurrent neural network, local data was gathered from the Santa Clara county website, with 325 columns of COVID cases,338 columns of hospitalization data, and over 335 columns of COVID death data since March 27, 2020, when the county started recording data. The RNN was trained on Pycharm with 50-100 epochs using 80% of the available data for training while using the remaining 20% of the existing data for testing. Data was downloaded up until February 26, 2021 and was then run to predict results forward up until March 26, 2021. This was then compared to the actual numbers that were subsequently recorded by the county. Overall the system is able to predict future covid trends that can be expected. Even though vaccinations are underway, this does not stop covid numbers but merely protects people from dangerous situations