We have leveraged our unique dataset of more than 40 million de-identified messages as well survey results from more than 1 million conversations to advance the use of AI/ML in delivering digital mental health services. This is one of the richest and most unique sources of insights for emerging mental health trends in Canada. We uses Natural Language Processing and other advanced analytics to extract unique insights from youth language as it relates to mental health topics. This analysis allows us to understand how youth talk about topics like climate anxiety, school stress, family relationships, and current events in their own words. Our goal is to expand our understanding of youth language as it relates to mental health to inform service design, prevention, and a general understanding of the complexities of youth language as it evolves over time. We will present results from our research model which can identify the top clinical issue in a conversation transcript with 90% accuracy. We will go into detail on use cases in developing automation for frontline staff and personalized recommendations for youth who are experiencing feelings for the first time. We will also share lessons learned and common hurdles in moving from applied research into product development. Attendees will walk away with a better understanding of how to set their org up for using AI/ML, including: - Developing use cases that are suited for AI/ML - Ensuring adoption of AI/ML accounts for bias - Using open-source technology (free) to develop use cases - The importance of investing in product development – not just research - Change management needs for frontline staff Our use of AI/ML has the potential to transform mental health service provision, offering scalable and impactful solutions to address youth mental health challenges.