This project will develop objective tools to predict treatment risk, early improvement, and how long benefits may last for multiple rapid-acting behavioral health treatments (e.g., ketamine, psilocybin) for multiple conditions (e.g., anxisty, depression, opioid use disorder). By combining clinical outcomes, speech patterns, and biological data collected over time across research studies and real-world care, the project aims to create practical measures that can support more personalized treatment and future FDA-ready mental health endpoints. Data collected under this award will be deidentified and housed in the ARPA-H EVIDENT Data Repository.