Quantitative behavioral coding of counseling treatments is time consuming, however rigorous in determining which aspects of our practices are efficacious. By using AI/ML, presenters share a 3 step process of enhancing productivity. First, we built a deep neural network classifier for one item to test ML algorithm. Next, we expanded the ML model to the rest of the coding protocol. Finally, we optimized the full ML coding model with > 8,000 audio files.