Enhancing Clinical Monitoring with Voice-Integrated AI
→ Product Designer
→ Mental health AI monitoring app
→ 2023 - 2024
About
Dasion is a NSF (National Science Foundation) funded AI health tech company that addresses healthcare issues with multimodal AI and voice biomarkers.
Its remote monitoring capability enables continuous assessment of patient mental health, issuing timely alerts to healthcare providers for necessary interventions.
Problem
Hypothesis
Focusing on what matters: we prioritize user engagement metrics to boost AI speed, accuracy, and trust.
I formulated design-focused hypotheses and prioritized them with cross-functional teams using an impact matrix to guide our development efforts effectively.
Vision
Monitoring post-diagnosis mental health treatment should ensure trust, reliability, and companionship.
Early iterations
Evolving our MVP: focusing on user engagement in recording for effective monitoring.
Designs across different service provider are not consistent & outdated.
MVP Framework
The Recording for Monitoring feature
The Recording for Monitoring feature is where our app’s core functionality comes to life. Users engage with the app through daily voice journals, designed to be brief yet effective, capturing essential data points for mental health monitoring.
Testing & Iteration
Users expressed a clear preference for recording their inputs rather than engaging in real-time conversation.
I led the testing with our internal user group, using their feedback to shape two key iterations of our app.
V1 - V2: The first round of feedback led us to separate the prompt and speaking functions, enhancing user focus and interface efficiency.
V2 - V3: In the next iteration, user preferences for flexible communication modes prompted the introduction of a chatbot that supports both typing and voice input.
Outcome
After we re-ran the testing, we were able to positively impact all metrics.
Onboarding Process
Following sign-up, users are presented with the Terms of Use in a clear, easy-to-understand format, ensuring transparency about data usage and user rights.
Recording for Monitoring
Users engage with the app through daily voice journals, designed to be brief yet effective, capturing essential data points for mental health monitoring.
Result Analysis
This feature provides critical insights into the user’s mental health trends and potential needs for intervention.
Impact
67%
improvement in efficiency.
The average time to begin recording was reduced from 30 seconds to just 10 seconds.
decrease in the number of attempts.
Clearer instructions and an improved user interface reduce user frustration and an increase in ease of use.
decreased the average time users spent recording their responses.
Refinements in the prompt clarity and enhancements decreased from 2 minutes to 1 minute.
3 to 1
50%