Case Study

The in-app AI assistant we built into UTrack for Utrack Inc: device location questions, settings changes and voice input inside the GPS tracking app.

UTrack: a phone showing an in-app AI assistant chat, with chat bubbles rising toward a glowing route map and pin beside a parked SUV
Cover

Gallery

UTrack App Store screen: the in-app AI Device Assistant answering the question "Is my car in a safe area?" next to a tracked car on the map

The challenge

A GPS tracker sold as a consumer service creates the same questions again and again. Where is my device right now? Why did I get that alert? How do I change this setting? The tracking app may already show a live map, alerts and route history, yet the answers to those questions usually go through email, a ticket form or a chat tool that knows nothing about the customer's device. That split is the gap. The app holds the live device context, and support sits outside it. Customers who just want a quick answer have to leave the product, describe their device to someone, and wait. Larger telematics suites have started to ship built-in assistants. For a device-plus-subscription tracking business running its own app, the question is how to give customers the same kind of help inside the app they already use.

Architecture

We built UTrack end to end for Utrack Inc, a US GPS tracking company: a Traccar backend we set up and customized, a Laravel web admin for the operations team, and a Flutter app for iOS and Android. Inside that app, we built an in-app AI assistant. This case study looks only at that assistant. The full UTrack platform, from the device inventory to subscriptions, is covered in the main UTrack case study on our site.

Solution

What the assistant gives UTrack customers: - Questions about the device's location, asked in plain language inside the app instead of through a separate support channel - Settings changes through the same conversation, so customers do not have to hunt through menus - Voice input, so a customer can ask instead of type - 24/7 in-app support, available whenever the customer opens the app The assistant sits beside the rest of the UTrack app: live location with speed, battery and signal, alerts in the app and by SMS and email, geofences with enter and exit alerts, live trip trails and route history playback. One Flutter codebase serves iOS and Android, so the assistant ships to both platforms together. Behind the app, the Traccar backend we set up and customized and the Laravel web admin run the devices, alerts and subscriptions for the whole service.

Results

What this changes for Utrack Inc and its customers: - Common device questions can be asked inside the tracking app, next to the live map, instead of in a separate support channel - Customers can change settings by asking, rather than finding the right screen - Voice input makes the app easier to use on the move - Support is available in the app around the clock - iOS and Android customers get the same assistant from one Flutter codebase Next step Running a GPS tracking app and want an in-app AI assistant that answers device questions inside the product? Talk to us at https://techmania.solutions

FlutterLaravelTraccar

Related project

UTrack

UTrack is a GPS tracking service by Utrack Inc (US) that we built end to end: a Traccar backend we set up and customized, a Laravel web admin for devices, alert features and subscriptions, and a Flutter app for iOS and Android with live tracking, alerts, geofences and route playback.

View project in portfolio
← All case studies