Empathetic Virtual Agents For Customer Service

Pilot #1

About this Pilot

Discover the future of smart home technology with our cutting-edge pilot project. Led by TID and in collaboration with esteemed partners, we are pioneering a privacy-preserving, decentralized Federated Learning framework to develop personalized, high-performance language models. This innovative approach ensures your data remains secure while enhancing the functionality of Telefónica’s leading products and services, from cognitive personal assistants to smart home management systems.


Harnessing the power of Federated Learning (FL), this pilot ensures that deep-learning models are trained in a decentralized, privacy-preserving manner, keeping data secure on local devices. TID will deploy and evaluate hierarchical FL models across diverse environments, from edge devices to telco infrastructures. This innovative approach allows Telefónica and its partners to create high-performance, personalized language models without compromising data privacy. These models will enhance a range of Telefónica services, including AURA personal assistants, Movistar TV recommendations, and smart home management systems.


Main Challenges

To successfully implement this pilot, we need to address several key challenges that will drive the innovation and efficiency of our privacy-preserving language model learning framework:



Enable continuous learning from user preferences, specific lexicon, and life routines through unsupervised algorithms, incorporating iterative design methodologies and user participation.


Ensure privacy in residential spaces by keeping private data within the home and only sharing models for global aggregation using FL.

Global Knowledge

Maintain global knowledge through model transfer techniques to address the cold-start problem and enable faster incorporation of advanced language models across Telefónica applications.

the safari park

Safari & Adventure Park

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Nullam quis sem lacus. Nunc venenatis ultricies sapien, vel placerat ante commodo id. Nullam turpis velit, porta quis tempor eu, posuere quis augue.

Explore other pilots

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo. 

Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo. luctus nec mattis, pulvinar dapibus leo.

Bias Detection In Language Models

Focusing on detecting social biases, the pilot utilizes cultural context to ensure language models in customer service scenarios respect societal norms and contribute to more equitable interactions.


Empathetic Virtual Agents For Customer Service

Enhancing customer service through retrieval-augmented LLMs, the pilot develops virtual agents that understand user goals, make informed API calls, and respond empathetically, aligning with user emotions and objectives.


AI-Enhanced Emergency Call Center Supervision

Transforming emergency health support in call centers, the pilot uses AI to supervise multimodal dialogues, assisting medical staff to provide timely and accurate advice through an innovative audio corpus and AI recommendations.


Recommended Resources

Explore this section to view Eloquence's key milestones and deliverables, access our latest publications, and stay updated with our newest technical releases. Don't miss out!


Small title