Project Abstract High attrition and low adherence rates continue to hinder the efficacy of digital addiction treatment programs. When individuals prematurely drop out of treatment programs, they are unable to realize the numerous benefits associated with completing an addiction treatment program, such as fewer relapses and higher levels of employment. Workit Health is developing a fully integrated chatbot to increase engagement in digital addiction treatment programs. The chatbot will be able to make customized, actionable recommendations to provide care administration and psychosocial support. The integration of this chatbot will also provide the opportunity for individuals to track their progress, identifying behavioral patterns, and incremental improvements. While other mental health chatbot technologies have been applied more broadly, offering support for individuals with anxiety and depression, our chatbot will be the first to specifically cater to individuals struggling with addiction. Moreover, our chatbot will be the first of its kind to be fully integrated into an addiction treatment program, adopting a hybrid human-first approach. Where other chatbots seek to replace certain roles of mental health providers, our chatbot seeks to identify areas that chatbots can support and augment mental health professionals. We expect this nuanced difference to lead to new insights into how medical providers can collaborate with a chatbot and new ways to utilize the technology. The long term goal of the project is to identify novel ways to engage members in online addiction treatment programs to increase retention. This project will explore how chatbots can be used to offer personalized and scalable interactions with individuals in addiction treatment programs. In Phase I, we seek to understand how we can engage members to utilize the chatbot. In Phase II, we will focus on identifying the best scenarios for a chatbot to respond to and how to persuade members to act on recommended items (such as scheduling an appointment with a counselor or attending a group therapy session). As leading health plans adopt value-based care models, it is imperative that addiction treatment providers can effectively communicate their impact through standard metrics such as retention, adherence, and appointment attendance. As member engagement increases, we expect to see improvements in these standard outcome metrics, allowing members to maximize the benefits that they receive from their addiction treatment program. This project will offer insights into best practices for other treatment providers looking to integrate a chatbot into mental health treatment programs.