Ivanna Sysak@Ivanka

We believe every severe crisis begins small. It could be as small as one grain of rice. "Invisible burdens"—small, overlooked stressors—can quietly escalate into major mental health challenges. We address these micro-crises with AI-driven support, offering tailored exercises for immediate relief. It bridges gaps in accessible care, empowering users to manage stress effectively. We’ve created a demo of a "text-to-CBT intervention" model. We use Llama 3.1 70B to generate questions that help clarify the user's situation. After this clarification, we provide the model with a description of the user's issue and ask it to generate a multi-step mental health exercise that may be helpful. The code for the exercise is then rendered in the user's browser using React. The prompts were developed with the support of a mental health professional. The advantage of the Llama 3.1 model here is its support for multiple languages, which means that, in the future, we will be able to fine-tune the solution to assist in Ukrainian. Our main audience is Ukrainians of all generations and identities, and our AI speaks their language. Why Ukrainians? Because we are them. Over 7 million Ukrainians need mental health support, with half facing anxiety, depression, or PTSD due to war. Access is limited, with just 1 psychiatrist per 20,000 people, and many professionals displaced or unable to work. Our service is free for users, generating revenue through collaborations with organizations, universities, and humanitarian groups to expand access. Unlike competitors, we offer free, quick, and personalized crisis relief, making it uniquely impactful. The addressable market ($100–200M) estimates 6–8M users seeking stress-related solutions in the target region, multiplied by the average revenue per user (ARPU). The Serviceable Available Market ($50–100M) narrows this to 1–2M users actively seeking immediate stress relief, focusing on our product's core use case.