Tech is transforming how businesses engage with their customers, and Mox is at the forefront.
As a rapidly growing digital bank, majority of our customer interactions happen through in-app chat. Real-time customer insights become crucial to make each conversation count.
Using machine learning and natural language processing, we developed a tool that can accurately analyse English and Cantonese text so we can understand the sentiment and the context behind every customer conversation.
This tool is a product of our concerted effort. We carefully considered the technical aspects and the end-user experience to select the right algorithms that improved accuracy. This approach allows us to explain how our model makes decisions, ensuring transparency and responses that are contextually relevant.
To create this model, we started by using open source language data to train a base sentiment scoring model with generic language processing capabilities. Then, we refined this model with internal data and expert knowledge to understand the unique nuance at Mox, leading to accurate customer sentiment scoring.
The model development is a testament to our ability to combine quality engineering, data science and teamwork to create the best possible customer experience.
The new homegrown model enables:
At Mox, we’re committed to making every interaction feel valued and leaving the customer heard. Technology empowers us to do this even better — that’s the Mox difference!