Voicebot Mobile App

Home
Case Studies
Voicebot Mobile App

About Project

To enhance and expand the mobile app of a bank in Vietnam, a new module was integrated, designed to provide a more convenient and modern user experience through voice commands. With the rapid development of artificial intelligence and voice recognition technology, banks are increasingly focusing on improving their services not only to meet current demands but also to offer customers a more advanced interactive experience.

This module offers a unique feature called 'Wake-up Service,' which allows users to activate the app with a simple greeting command like 'Hi, Bank.' This feature is highly convenient as it enables users to open the app and start interacting without touching their phones or unlocking the screen. It creates a seamless connection between the user and the bank, enhancing user interaction and intimacy, and providing a feeling of being continuously managed and supported.

Project Information

industry

Finance and Banking

Project Type

Mobile App

Country

Vietnam

Team Size

10+ members

Tech Stack

Swift, Kotlin

Project Challenge

Voice Recognition Accuracy: Accuracy is crucial in financial transactions, and it can be challenging to accurately recognize and interpret various Vietnamese dialects, accents, and speech patterns.
Wake-up Command Responsiveness: Ensuring the app reliably responds to the 'Hi, Bank' wake-up command without false positives or ignoring actual attempts.
Performance on Various Devices: Ensuring consistent performance across multiple devices with varying hardware capabilities.
Handling Complex Transactions: Managing complex transactions or sequences of transactions via voice commands without confusing the user.

Project Results

Utilize advanced Natural Language Processing (NLP) tools and continuously train the voice recognition system with diverse datasets to improve accuracy. Integrate alternative mechanisms for confirming critical commands via text or touch input.
Optimize the voice activation module with noise-canceling technology and allow users to customize wake-up commands. Also, regularly update recognition algorithms to adapt to different environments and usage scenarios.
Break down complex transactions into small, verifiable steps and provide clear, simple guidance. Allow users to review or cancel commands before execution. Optimize the voice bot for both iOS (Swift) and Android (Kotlin) platforms. Conduct extensive testing on various devices to identify and resolve performance issues.