DeepKeys

DeepKeys

Mental Health Monitoring

81 followers

Visit website
DeepKeys is a mental health monitoring app. It functions like a heart-rate monitor for the mind, helping users track their mental well-being. The app allows for social sharing, enabling users to follow the mental health journeys of their friends and family.
DeepKeys gallery image
DeepKeys gallery image
DeepKeys gallery image
DeepKeys gallery image
DeepKeys gallery image
DeepKeys gallery image
DeepKeys gallery image
DeepKeys gallery image
DeepKeys gallery image
Free
Launch tags:
Productivity•Artificial Intelligence•Quantified Self
Launch Team

What do you think? …

Matthew Olsen
Maker
📌
Hello Product Hunt! We're thrilled to introduce DeepKeys, a mental health monitoring app that tracks your mental well-being like a heart-rate monitor for the mind. Perfect for self-improvement enthusiasts and concerned family members. Key Features: • Mental Health Tracking: Monitor and understand your mental well-being over time. • Social Sharing: Share and follow mental health journeys with friends and family. • Actionable Insights: Get insights to support personal growth and well-being. We created DeepKeys to provide an accessible tool for mental health tracking. We'd love your feedback to help us improve and grow.
Christofer Huber
@matthew_olsen Hey Matthew, 2 questions: 1. Where are you getting the data from? I assume you use the Apple Health data? 2. How are you doing generating these summaries, do you use a LLM for that or do you have your own algorithm? Congrats on the launch 🚀
Matthew Olsen
Maker
@crebuh 1. We currently use input monitoring with some post-processing for privacy & security as the primary data source. We then run an LLM on their desktop by default (or in the cloud if they choose that in the settings) & discard the raw data after it's been analyzed (see FAQ: https://www.deepkeys.ai/faq). 2. We use an LLM to summarize the data with some preprocessing involved.
Kyrylo Silin
Hey Matthew, How exactly does the app measure and quantify mental well-being? I'm also curious about the social sharing aspect. How do you balance transparency with privacy concerns? It would be interesting to hear about any insights you've gained from aggregated user data so far. Congrats on the launch!
Matthew Olsen
Maker
@kyrylosilin Thank you for the questions & the congratulations. 1. We run text classification for a variety of metrics to estimate a user's state based on their inputs. Over longer intervals (say 15-min to 1-hr) the quality of the predictions seems to get better and there is less noise. 2. By default, the user's accounts are set to private. The users will have to accept follow requests from other users before the data can be shared. In the future, if users request it we can add features that limit which data, graphs, and summaries are shared.