As a part of the expansion of services in one of our banking clients in Hong Kong, we introduced a new set of capabilities to our existing Text-to-Speech (TTS) product which allowed for more of a dynamic script. Historically, what we have enabled is a user to dial into a number and then listen to a static list of risks associated with a product and its risk. While that may have worked, we think it's time for a change. Improving UI, Improving ProductivityPresenting an improved TTS product, we have built a user interface (UI) in which users can type in key information about a loan. These inputs are then fed into our system, which would dynamically create a script (which includes the data fed into the system) to convey the terms and conditions to the user.Typically, this would have otherwise taken some time for users to manually put together. However, this has now been digitized so that the scripts are built pertaining to the product. Not only this, this saves the RM’s time in reading out the terms and conditions - giving RMs more time to focus on other more productive tasks while the user listens to the terms and conditions.While it's still in the implementation phase, the concept has been tested and previous iterations of the product (albeit without the dynamic script generation abilities) have successfully been implemented at the bank, and is now being widely adopted.What's Next?We are constantly improving the abilities of the TTS abilities. For example, we have found that the product sounded too robotic and machine-like. We're working hard to review ways in which we can make the sound to be more natural and human-like.