Web Virtual Agent Framework

Web Virtual Agent Framework

Back to projects

An interactive conversational interface that personifies AI with a real-time animated avatar, complemented by an optional LESCO sign language translation module for video calls.

This solution takes digital interaction to the next level, giving AI a face and a voice to create dynamic, fluid, and deeply human conversations. Instead of interacting with a traditional text chatbot, users can communicate naturally via voice or text with an animated avatar that reacts and engages visually in real time.

The experience stands out for its high level of expressiveness and instant response. The avatar comes to life by adapting to the rhythm of the conversation: breathing subtly when idle, actively listening when spoken to, displaying a visual thinking state while processing information, and vocalizing with precise lip-syncing when delivering a natural-sounding spoken response.

As an additional module and accessibility use case, the platform also integrates a Google Meet Chrome extension. This complementary tool captures live subtitles during virtual meetings and translates them into Costa Rican Sign Language (LESCO) through the avatar, demonstrating the tremendous versatility of this architecture across customer support, online education, and digital inclusion.