mCS-LM, advanced customer service ecosystem

This initiative seeks to investigate and integrate technologies that allow the efficient processing of text, audio and image, using Retrieval Augmented Generation (RAG) systems to minimize hallucinations and improve the accuracy of responses, training the models with information specific to the client's domain.

The project envisions the creation of modules to centralize incident management (claims, returns, warranties), prioritizing them through automatic classification and emotion analysis. In addition, autonomous virtual assistants will be developed to resolve frequently asked questions and assist human agents. All of this will be complemented by a multichannel analytics system that will offer an interactive dashboard to monitor performance in real time, facilitating data-driven decision-making. Finally, all these components will be integrated into a framework validated in real-world use cases.