SUMMARY

CuraJOY’s AI coaches provide behavioral health support to families, but they need to improve accuracy, expand their content library, and involve users in the training process. My role was to design a platform that empowers clinicians, agents, and community members to collaboratively train and refine the AI, ensuring it meets diverse user needs.


YEAR

2025

ROLE

UX Designer

CONTRIBUTION

Research · Ideation · Wireframes


The Challenge:

CuraJOY’s AI coaches were effective but needed continuous improvement to better serve diverse families. Key challenges included:

  1. Limited AI Training Tools: Clinicians and agents lacked efficient ways to fine-tune the AI or provide feedback.

  2. Lack of Community Involvement: Community members wanted to contribute but had no accessible way to do so.

  3. Privacy Concerns: Clinicians needed PHI (Protected Health Information) redaction to ensure compliance and user trust.

Why It Matters:

Improving the AI coaches ensures families receive accurate, personalized wellness support. Involving communities fosters collaboration and trust, while robust privacy safeguards protect sensitive user data.


Research Goals:

  • Understand the needs of clinicians, agents, and community members.

  • Identify pain points in existing AI training workflows.

  • Explore opportunities for community-driven collaboration.

Research Methods:

  • Interviews: Conducted 10 interviews with clinicians, agents, and community members to understand their needs and pain points.

  • Surveys: Distributed a survey to 50+ users to gather quantitative data on feature preferences.

  • Competitor Analysis: Analyzed platforms like Woebot Health, Replika, and Hugging Face to identify best practices and gaps.

RESEARCH & DISCOVERY