Slingshot AI
Slingshot's mission is to develop first-of-its-class AI models for mental health. I worked with the founders and engineering team to design and launch their first consumer product, Ash: a voice-based AI therapist that brings mental health to anyone with a mobile phone. The app uses natural language to simulate therapy sessions with a trained professional over voice and text.
Services
Date
Feb - Apr '24 (10 weeks)
Onboarding
The goal of the app's onboarding was to collect the relevant data for Ash to understand how to best help each user. After several explorations, we landed on a small survey that quickly jumped users into the first session with their coach.
Session duration
The team wanted to encourage 15 minute sessions with Ash and a natural way to wrap up each session. However, a countdown clock felt stressful in this context. I explored a few ways to represent the passage of time without relying on a countdown clock.
Modalities
The Slingshot team was ahead of its time exploring multi-modalities when interacting with LLMs. The goal was to support text, tap-to-speak, and hands-free audio to encourage natural and convenient ways to interact with your AI coach.
Using a single button to communicate 3 states: idle, listening, and thinking
Early exploration of a more stark speaking (light) and listening (dark) states
Demo of switching between 3 modalities: tap to speak, hands-free audio, and text
Website
For Ash's website, we focused on finding imagery that conveyed the feeling of using the product. I explored a few different illustration styles but ultimately landed on a more abstract visual treatment.
