MobileHealthcare2023

Talk & Relax

Mental wellness mobile app โ€” React Native + Node.js backend + full admin panel.

The Challenge

Mental wellness apps occupy a crowded market, but most fall into two categories: meditation-only tools with no human connection, or full teletherapy platforms with clinical overhead that puts them out of reach for everyday stress management. The client identified a gap for a "soft support" layer โ€” guided relaxation content combined with moderated peer interaction โ€” that didn't require licensed therapist sessions but was still evidence-informed and professionally curated.

The technical requirements were demanding: audio-visual guided sessions needed to stream reliably on 4G connections across South and Southeast Asia (where average bandwidth is lower and more variable than Western markets), peer chat rooms required content moderation at scale, and user progress data (streaks, session history, mood logs) needed to be portable across device upgrades โ€” a common churn driver in the category.

What We Built

The mobile app is built in React Native with Expo, using react-native-track-player for background audio playback (guided sessions continue playing when the screen locks, a non-negotiable UX requirement). Guided session content โ€” audio files and accompanying animation scripts โ€” is stored in Supabase Storage and served via a CDN, with adaptive bitrate selection based on the device's reported connection quality at session start.

The backend is Node.js/Express with a strict Controller โ†’ Service โ†’ Provider layer pattern, deployed on Railway with PostgreSQL (Prisma). Socket.io powers the peer support chat rooms with message persistence and a real-time typing indicator. Content moderation in the chat rooms runs a two-pass filter: a keyword blocklist for immediate suppression, plus a daily async review queue where flagged messages are reviewed by the moderation team via the admin panel.

The admin panel โ€” built in React with Recharts โ€” gives the content and clinical teams full visibility into session completion rates, mood trend aggregates, and peer room activity, segmented by cohort. Push notifications for daily session reminders and streak milestones are sent via Expo's push notification service (with a Firebase and APNs backend), with send-time optimisation that respects each user's historical engagement window.

The Outcome

The platform grew from zero to 10,000 monthly active users in six months post-launch, with a Day-30 retention rate of 38% โ€” well above the 20โ€“25% benchmark for wellness apps in the same category. The App Store rating stabilised at 4.7 stars, with reviews consistently praising the audio quality of guided sessions and the non-judgmental tone of the peer rooms.

Infrastructure costs scaled linearly rather than exponentially: the Socket.io horizontal scaling architecture (Redis adapter + stateless API nodes) meant that adding capacity during growth spikes was a one-command Railway deployment rather than a manual server migration. The content team used the admin analytics to identify that 7-minute sessions had 2ร— the completion rate of 15-minute ones โ€” leading to a content strategy pivot that further improved engagement metrics.

Building something similar?

We'd love to hear about your project.

Let's talk โ†’