AI Mental Health Companion
Live
Problem
Users needing support outside clinical hours had limited immediate options, and providers lacked a scalable way to maintain monitored engagement with users between sessions.
Approach
Researched mental health support best practices before designing the conversational flow, to ground prompt design in appropriate, safe interaction patterns.
Outcome
Delivered to client.
Status
Live
Role
Defined the conversational flow and assisted in developing prompt engineering architecture. Owned the project roadmap, task tickets, and design reviews.
Target Audience
Patients in active therapy, Out of hours care seekers
Team Size
4
Project Duration
Not specified
Key Skills
Prompt engineeringconversational designUX
AI powered conversational companion providing on-demand mental health support with real-time sentiment monitoring, extending accessible emotional support beyond standard clinical hours. Users needing support outside clinical hours had limited immediate options, and providers lacked a scalable way to maintain monitored engagement with users between sessions.
My Approach
Researched mental health support best practices before designing the conversational flow, to ground prompt design in appropriate, safe interaction patterns.
Key Challenges
Designing prompt responses that felt empathetic and human rather than robotic or scripted, a core requirement for a mental health support context.
Defining safe clinical boundaries for what the AI should and shouldn't handle, ensuring the companion supported users appropriately without overstepping into areas requiring licensed clinical judgment.
Delivered to client. Adoption and usage impact not yet measured.