The document summarizes the results of two projects analyzing job posting and unemployment data: 1. A model was created to classify good and bad job postings, achieving 79% accuracy. It found the top parameters for successful postings were character count, word count, and estimated salary. 2. A model was developed to forecast monthly unemployment rates three months ahead based on state and month. Using random forest and decision tree algorithms, it achieved 73-76% accuracy. Future improvements could include adding lagged data and more granular unemployment features.