Gemma Hall

UX + AI CONTENT | MULTIMEDIA SERVICES

TARA AI

As part of a team project, we developed the concept of a TARA AI: Mental Health Triaging Assistant. The concept is aimed at reducing wait times and easing the discomfort of discussing sensitive issues with a human. The AI assistant is designed to assess consumers' therapy needs and match them with the most appropriate therapist based on their personal requirements and mental health concerns.

The mock-up UI was designed in Figma and tested using a "wizard of oz" approach, where participants interacted with the AI interface, but human responses were provided behind the scenes. Visualizations supported the experience, ensuring the interface felt dynamic and responsive.

To evaluate its effectiveness, we conducted an A/B test, comparing the AI version to contemporary online mental health testing checklists, such as the DSM-5. The results were clear: all participants expressed a strong preference for the AI-driven version, citing that it felt more personal and engaging than traditional checklists, which they found to be overly clinical and impersonal. This feedback validated the assistant's potential to improve the therapy triage process in a way that is both sensitive and efficient.

View TARA AI prototype

9/10

participants suggested interest in the concept of an AI triage assistant for mental health related concerns.

10/10

participants indicated a preference for the model in contrast to current online self-assessment triaging seen in online-provided mental health support platforms.

of those surveyed and interviewed demonstrated interest in AI-driven mental heath support tools.

7/10

UX PROCESSES

User Research & Persona Development:

Conducted user interviews and surveys to identify the discomfort and frustrations individuals face when seeking mental health support, specifically around long wait times and the challenge of discussing sensitive issues with a human therapist.
Developed user personas representing key demographics, including young adults, busy professionals, and individuals new to therapy.

User Journey Mapping:

Mapped out the user journey from initial engagement with the TARA AI assistant to being matched with an appropriate therapist based on the user’s specific mental health concerns.
Identified user needs, motivations, and pain points at each stage of the triage process, from assessment to therapist matching.

Wireframes & Prototyping:

Designed low-fidelity wireframes in Figma to lay out the user interface for seamless navigation and interaction with the AI assistant.
Created high-fidelity interactive prototypes and tested them using a “wizard of oz” approach, where users interacted with the AI interface while responses were provided by a human behind the scenes.

Usability Testing:

Conducted A/B testing to compare the AI-driven experience with traditional mental health testing tools, such as the DSM-5 checklists.
Gathered feedback from participants, who overwhelmingly preferred the AI assistant for its personalised and engaging approach, which they found to be less clinical and more comfortable than traditional methods.