Lax | Devpost

Do some activities
Earn points
Get in touch with community
Discover new things

Inspiration

Our mental health and wellbeing are currently threatened as we face lockdown; many of us are working from home, trying so hard to cope with the current situation which could easily lead to burnout and anxiety; According to the WHO “Burn-out is a syndrome resulting from chronic workplace stress that has not been successfully managed.

What it does

Lax is a mobile app that helps users to deal with stress by changing daily working routine and activities and introducing the mindful practice to their working hours.

1) The everyday healthy activities tab aims to better the lifestyle of employees wherever they are. There are various healthy options and reminders throughout the working hours to do something healthy such as drinking enough water and practicing a few minutes meditation exercise, and unlocking the badges as they achieve their monthly target; in order to be rewarded with a badge that motivates our user to have discounts on our various services.

2) Community tab where you can invite your friend for a healthy activity to do during the working day; and get the discounts as you both successfully complete the challenge from our key partner( doctors, gyms, yoga instructors, etc)

3) Stress detecting: in the early-stage model, through a chatbot; measuring everyday stress levels through text messages and measuring the stress levels of our user.

How I built it

Stress Detecting a super AI model that predicts stress at early-stage, based on user response measuring the daily stress levels of our user.

We divided our ML models into three main parts (stages). In the first part, we use Text data (collected from tweets) to measure stress from spoken language. Every single text has two scales. One for stresses the second for Relaxation. Then we combine between the two scales to detect users under stress.

Next stage we would like to incorporate Heart Rate Variability (HRV). Based on thirty-five different features the model predicts the stressor with a 99.25% accuracy. The third part uses a person’s thermal imaging to detect stress levels. Based on breathing patterns as temperature changes and breathing rate. Finally, we use the output from the three previous models to make an ROI from our users.