Connected mobile product on AWS IoT Core

Confidentiality: this page describes architecture and outcomes only.

The problem

Users needed to monitor and control home appliances — lighting, thermostats, security devices — in real time from one app on Android and iOS, with scheduling and automation.

The hard parts were latency on device state updates, secure device communication, and an architecture that would let new device types be added without rewriting the app.

What I designed and built

1. Real-time device communication over MQTT with AWS IoT Core for device management and state synchronisation, with background sync and local caching so the UI stayed responsive on poor networks.
2. Security. TLS for all device traffic, token-based authentication, and safe handling of device credentials on the phone.
3. Extensible device model. A device-capability abstraction, so a new appliance type is a configuration and a small adapter rather than a new set of screens.
4. Automation features. Scheduling and personalised automation rules, which became the main driver of daily use.
5. Release management. Unit and UI test suites, store submissions, and post-launch monitoring.

Stack

Flutter / Dart · MQTT · AWS IoT Core · Google Cloud IoT · TLS and token-based authentication · Google Play and App Store release

Outcomes

• 40 percent reduction in latency after caching and bandwidth optimisation.
• 4.8-star rating on both stores.
• 30 percent increase in daily active users within three months of launch, after the automation features shipped.
• New device types added without architectural changes — something the client called out specifically.

Related mobile work

• Retail banking app. Kotlin, Swift and Flutter; biometric authentication, AES-256, secure token handling; 50 percent faster load times; regulatory compliance testing; shipped to both stores.
• Enterprise employee apps. HR self-service, leave workflows, live bus tracking with Google Maps and push notifications; Flutter, Node.js, PostgreSQL, Firebase.