Learn how Drexel University improved turn-around time for sending decision letters by 75%. Using the unsupported test score feature, we created an upload process for a predictive model score. A numeric score between 100 and 999 is calculated based on several characteristics. A comma delimited file is created which is then loaded into Banner using SRTLOAD. Once the score is pushed to SOATEST the automated decision assignment process (SARBDSN) is run to assign a decision. Rules built on the decision rules table (SAADCSN) look at a range of scores and assign offer, deny or waitlist decisions depending on the model score. Based on the decision, a letter is automatically initiated depending on material rules built on SOAMATL. The entire process from creating the model score upload file to mailing the decision letters is done within 48 hours. Using this model, we have been able to increase productivity and drastically reduce the need to review applications, update decision codes and order decision letters. This presentation will cover: Setting up the Model score to run SRTLOAD, SRRSRIN and SRRPREL Setting up Decision Rules on SAADCSN Running the SARBDSN process to assign the decision and building the material rules to assign a letter.