Post by Josephine Loiret-Bernal

Just built my first Generative AI web app and it tackles something deeply misunderstood: OCD. Hi, I’m Josephine, and I’m excited to share a project I recently completed that combines machine learning, generative AI, and mental health care. Why OCD? Obsessive-Compulsive Disorder affects millions, yet it’s often reduced to quirks or stereotypes. It’s not about being neat or clean—it’s a complex, often debilitating condition. Many struggle for years without proper diagnosis or treatment. The Idea I created a web app designed to assist psychiatrists in generating tailored treatment recommendations for OCD patients, not for self-diagnosis or use by patients themselves. Implementation: 1 - Random Forest Predictive Model trained on a cleaned and encoded Kaggle data set (1,114 patients, 6 key features) to classify the appropriate drug group (SSRI, SNRI, Benzodiazepine). 2 - OpenAI’s GPT-4o then takes the predicted drug group and patient profile to recommend more specific treatments, with clear disclaimers about professional validation. 3 - Flask + HTML Web App to tie it all together. Users (psychiatrists) input patient data, get back recommendations via a clean interface. Why it matters: Too often, finding the right treatment is trial-and-error. This app aims to support clinicians by combining evidence-based predictions with generative reasoning which saves time, offers ideas, and reduces uncertainty. #AI #GenerativeAI #MentalHealthTech #OCD #MachineLearning #Python #OpenAI #HealthcareInnovation #WomenInTech

Video Content