SDAIA and SCAI Unveil "SauTech" Speech-to-text Software

General News
Published 14 Sep 2022
Share the article
SDAIA and SCAI Unveil "SauTech" Speech-to-text Software

Riyadh, SPA:
The Saudi Data and Artificial Intelligence Authority (SDAIA) and its National Center for Artificial Intelligence (NCAI) today unveiled the brand-new SauTech system, an AI-driven solution that converts speech into text using advanced speech-recognition techniques. Developed in collaboration with the Saudi Company for Artificial Intelligence (SCAI), a fully owned subsidiary of the Public Investment Fund (PIF), the new system was announced during the second edition of the Global AI Summit at King Abdulaziz Conference Center in Riyadh.
SauTech utilizes cutting-edge speech recognition algorithms to process spoken language. The software supports audio files of various sizes to enable real-time speech-to-text conversion across different dialects thanks to its precise ability to recognize and identify local Saudi dialects. In addition to outperforming all current applications, the SauTech system is set to be upgraded to recognize more Arabic dialects.
President of SDAIA Dr. Abdullah Alghamdi thanked the Custodian of the Two Holy Mosques and His Royal Highness Crown Prince Mohammed bin Salman for their tireless support of the Kingdom’s AI sector. He commented: “The innovative SauTech system has been developed entirely by Saudi AI scientists and engineers within the Kingdom. This solution is one of many initiatives from SDAIA and NCAI that are set to shape the future of the AI sector and harness the power of data to serve the national economy and Saudi society. We will continue to seek out new ways to enhance the Kingdom’s role in developing, innovating, and adopting these technologies and cement the Kingdom’s leadership among data and AI-based economies.”
NCAI’s Chief Executive Officer Dr. Majid Altuwaijri commented: “We are proud to announce SauTech as the first product from NCAI in collaboration with our partners at SCAI. The new system has been extensively trained using huge amounts of audio data capturing diverse dialects and accents to ensure the highest levels of accuracy and performance. The system places privacy as a priority, with all user data stored on servers within the Kingdom to ensure the utmost level of security. This innovative solution adds tremendous value and further strengthens the Kingdom’s AI capabilities.”
Eng. Ayman AlRashed, Chief Executive Officer of SCAI, said: “We are proud to partner with SDAIA for the development of the innovative SauTech system, which delivers outstanding accuracy and recognizes local dialects in a way that outperforms all current applications. This technology will create new opportunities to deploy AI across many more applications, in line with PIF’s strategy of promoting innovation, building capacity, and enabling promising sectors to achieve the goals of Vision 2030.”
Currently available through the ‘software as a service’ model, SauTech system will help create new opportunities to deploy AI across many more applications including assistive technologies and Interactive Voice Response systems (IVR) used in emergency services, healthcare, and banking. The technology also enables automated captioning of media, call center analytics, automated meeting transcriptions, Conversational Agents, and voice-activated devices.
Results of a benchmarking analysis carried out by aiXplain indicate that SauTech outperforms the next best system by approximately 8% and 9% on Modern Standard Arabic and Saudi dialects respectively.
The Saudi Data and Artificial Intelligence Authority and SCAI have signed MoUs with SAUDIA and stc to harness artificial intelligence technologies and language processing solutions, including SauTech, to improve the two partners’ customer experience.
Work is underway to add more Arabic dialects to SauTech system, which is being exhibited for the first time at the SCAI pavilion during the Global AI Summit in Riyadh.

Last Modified Date: 14/09/2022 - 1:34 PM Saudi Arabia Time


Welcome to our search page

What are you looking for?