THE BRIDGE

tag QA Engine

Japan’s Studio Ousia unveils bot solution leveraging Quiz Bowl-winning technologies

SHARE:

See the original story in Japanese. Tokyo-based Studio Ousia revealed last week that it had completed an AI (artificial intelligence) inquiry response system using deep learning named QA Engine and that the system was adopted to the auto-answering system of Japan’s cloud accounting system Freee as its first client. The system supports the questions-and-answers style and can hand over questions to human operators when users are dissatisfied with the answers. Freee expects an improvement in customer satisfaction and the efficiency of helpdesk works through the QA Engine adoption. Studio Ousia won an AI quiz competition in Quiz Bowl style at an international conference focused on natural language processing NAACL (North American Chapter of the Association for Computational Linguistics) held in San Diego last June. QA Engine was developed leveraging the competition-winning technologies. The firm plans to launch QA Engine as an API (application program interface) service of elemental technologies applicable to answer support system in call centers, human resource matching system or chatbot. See also: Japan’s Studio Ousia wins global NEEL Challenge competition with Entity Linking technology Japan’s Studio Ousia to launch new affiliate solution for bloggers Japan-based Phroni introduces smart add-ons for Dolphin and Sleipnir mobile browsers As recent…

Image credit: 123RF

See the original story in Japanese.

Tokyo-based Studio Ousia revealed last week that it had completed an AI (artificial intelligence) inquiry response system using deep learning named QA Engine and that the system was adopted to the auto-answering system of Japan’s cloud accounting system Freee as its first client. The system supports the questions-and-answers style and can hand over questions to human operators when users are dissatisfied with the answers. Freee expects an improvement in customer satisfaction and the efficiency of helpdesk works through the QA Engine adoption.

Studio Ousia won an AI quiz competition in Quiz Bowl style at an international conference focused on natural language processing NAACL (North American Chapter of the Association for Computational Linguistics) held in San Diego last June. QA Engine was developed leveraging the competition-winning technologies. The firm plans to launch QA Engine as an API (application program interface) service of elemental technologies applicable to answer support system in call centers, human resource matching system or chatbot.

See also:

As recent topics in this field, Japanese online shopping mall Lohaco’s inquiry response system named Manami-san, which was developed based on a chatbot Caiwa provided by Exiis-Lab, received attention by becoming able to cover one-third of all inquiries to its customer support service. AI developer AL+ (pronounced as “alts”) developed an API Rewritable Memory-based Retrieval (RMR) as a core technology of chatbot and launched a bot development environment AL+ BotFramework supporting LINE or Twitter too. Japan’s BPO (business process outsourcing) / call center service major TransCosmos (TSE:9715) invested in New York-based bot development environment startup Reply.ai this month, and announced that the firm is going to localize the platform for the Japanese market while focusing on introduction to customer support services of the Japanese enterprises.

Studio Ousia was founded through the Keio University SFC (Shonan Fujisawa Campus)’s incubation program, and launched a keyword autolink plugin for smartphone / browser named Phroni in 2012. The firm raised 70 million yen (about $610,000) from Nissei Capital in March of the same year in its series A round, and 100 million yen (about $870,000) from Tokyo-based system developer NID (TSE:2349) in August 2014.

Studio Ousia CEO Ikuya Yamada explains the base technologies of QA Engine at NAACL
Image credit: Jordan Boyd-Graber

Translated by Taijiro Takeda
Edited by “Tex” Pomeroy