PC App Architecture

The Tomato.ai solution has two parts. A PC App that is used to deliver the solution, and a Web Console that is used to monitor usage of the solution. Below we cover the architecture for each.

Works with Any Communications App

The Tomato.ai solution works with any softphone or communications application. It uses a virtual microphone to pick up the voice from the microphone and then enhances it. It only takes a few minutes to install on a Windows PC. There are two features offered, each served differently, as detailed below.

Accent Softening Architecture

The Accent Softening solution picks up the voice using a virtual microphone and sends the voice for enhancement to Google Cloud, in region, close to the agent to limit network latency. The enhanced voice is then sent back to the PC and then out as normal to the communications or softphone application. The process ensures minimal delay, providing a seamless experience for both the agent and the customer.

The architecture is designed with high scalability and reliability, leveraging Google Cloud’s robust infrastructure. Each session dynamically connects to the nearest data center, optimizing voice clarity while maintaining low latency.

PC App data is sent end-to-end encrypted, ensuring that all communications remain secure. No data is stored on the cloud, aligning with stringent data privacy requirements. This design prioritizes user privacy and complies with international data security standards. For more information, see the next section covering security certificates.

Noise Cancellation Architecture

The Noise Cancellation solution is on device (i.e. running on the PC) and so the data does not leave the PC. This AI model is a lighter weight one and so can easily fit on lower end PCs. For more information see the next section covering security certificates.

PC App Installation

The current install is an .exe file. A Microsoft Software Installer (MSI) option is coming soon.

For details on the minimum PC required to run the PC App visit the Install the PC App (Beta) page where there are also details on how to setup the solution on a PC.