Accelerat.AI

Accelerating digital transformation in Portugal

Accelerat.AI

PROJECT OVERVIEW

Accelerat.ai aims to create disruptive solutions based on Conversational Artificial Intelligence (AI) Agents and CCaaS (Contact Center as a Service), which will accelerate more efficient interaction between public/private entities and customers/citizens. This will create new, more sustainable digital business models for a Future Economy. The goal is to pioneer a platform for cognitive services in European Portuguese (across various accents and age groups) that will completely transform the user experience in interacting with products or services. This innovative platform will be more inclusive, always available (24/7), and customized by industry, anticipating the resolution of about 80% of support cases. The consortium will invite all Portuguese, from different regions of the country, to contribute – with their voice and their knowledge as native speakers – to the training of AI models, creating disruptive services for this new Digital Era in Portugal and Europe.

The Accelerat.AI project is funded by the European Union through Next Generation EU as part of the Plano de Recuperação e Resiliência (PRR). More information regarding the project is available here.

OUR INVOLVEMENT

We are developing a conversational AI assistant that can understand and interact in languages often overlooked by the world’s largest tech companies. Our first aim is to automate 80% of customer support resolutions in Portugal, which will lead to significant cost savings for both the public and private sectors. This is a step towards our broader vision: to make high-quality Al technology accessible to a wider variety of language speakers worldwide.

At INESC-ID, R&D is led by Alberto Abad, a member of the Human Language Technologies (HLT) group and Associate Professor at IST.

R&D ACTIVITIES

  1. Development and training of a Conversational Al system that integrates several components: Automatic Speech Recognition (ASR), Text-to-Speech (TTS), Natural Language Understanding (NLU), Natural Language Generation (NLG), and Dialogue Manager (DM);
  2. Collection of crowdsourced data to feed the conversational Al models (focused on European Portuguese);
  3. Development of automatic data collection quality tools;
  4. Development of “Plug&Play” agnostic orchestrator and future integration guidelines, in line with existing market standards, that will allow clients to pick and choose whatever module they need to use and integrate it with possible existent solutions;
  5. Creation of Speech and Natural Language Processing toolkits with data collection that will mitigate the barriers for the entrance of new players that pretends to develop speech models for European Portuguese;
  6. Launch a Personal Identifiable Information redaction toolkit that will enable a compliant use of client data to treat new models;
  7. Assessment of responsible and ethical Al compliance that will perform accessibility evaluation of Al conversational systems over variances such as disabilities, gender, age or accent;
  8. Ensure that Virtual Agent, Automation Designer and Knowledge Management Al will decrease the needed effort to configure and maintain the conversational pipeline with the client;
  9. Allowing voice gateway to integrate with any Contact Center Platform;
  10. Ensure the integration of the developed models in the Contact Center Platform, allowing direct client assistance in support scenarios, by solving up to 80% of cases without the need for a human agent,
  11. Execution of dissemination and communication activities to raise awareness of the Agenda’s solution among the scientific community, target groups, and relevant stakeholders/sectors;

The integrated solution to be developed in Accelerat.ai can be applied in a wide range of sectors, being either public (Healthcare, Social Security, Administrative Services) or private (Retail, Insurance, Banking, Telco, Energy). The solution should allow entities to provide better remote customer satisfaction, and to lower their operation costs, since AI can provide 337% R0I using virtual agents across their organization. Accelerat.ai will foster Portugal as an AI Hub in Europe, attracting global talent, while pushing the country to become a case study of digitalization leveraged by AI.

Location

INESC-ID Lisboa

Rua Alves Redol, 9

1000-029, Lisboa, Portugal

Contact

alberto.abad@inesc-id.pt

PROJECT FUNDING

The Accelerat.AI project is funded by the European Union through Next Generation EU as part of the Plano de Recuperação e Resiliência (PRR). More information regarding the project is available here.

© 2024, INESC-ID. All rights reserved

});