Introduction

For this project, I was hired by a company, Clevvur, to develop two products: a phone bot and an email bot. These bots automatically respond to customer inquiries for the stores that use them.

The bots work in tandem with a dashboard that allows customization, such as modifying the knowledge base they use.

Voicebot

In the following interface, the user can add questions that might be asked and set corresponding answers. Using NLP, the system will intelligently identify a user’s question and select the appropriate answer. The user can also preview the audio generated for all the answers.

image.png

And in the following page, the user can see the usage analytics for the voicebot:

image.png

Here, the user can define the various managers to whom the receptionist can redirect the customer.

image.png

Voicebot Demonstration

haaa ‐ Réalisée avec Clipchamp(1)(1).mp4

Email bot

The email bot works in a similar fashion, allowing the user to define the knowledge base and specify the managers the bot can CC.

image.png

More information can be found at https://clevvur.io/

Conclusion

The creation of these bots, particularly the phone bot, was a technical challenge. Optimizing response time was crucial to make the call experience enjoyable, so I worked extensively on this optimization to process the customer's voice, transcribe it, and deduce the appropriate response as quickly as possible. I am very pleased with the result, as the response time achieved is faster than when using tools from competing companies like Vapi.