NTT Resonant Inc. is using NVIDIA Jarvis, a framework for conversational AI, with NTT Resonant’s conversational AI engine to build a demo of a freeform restaurant reservation system.
The demo enables a full reservation service, from searching for a restaurant to reserving a table, using freeform spoken conversation with an AI agent. The demo was unveiled during “NVIDIA GTC21” held from April 12 to 16, 2021.
NTT Resonant’s investment in AI technology and R&D has been recognized by international conferences such as the Association for the Advancement of Artificial Intelligence (AAAI) and the International Joint Conference on Artificial Intelligence (IJCAI). The goo AI x DESIGN solution uses goo AI technology for AI services, and a new task force at a goo lab, an R&D team specialized in data analysis and cutting-edge machine learning, is leveraging NVIDIA AI to build this demo.
This restaurant reservation system was developed using conversational intent comprehension technology, ConvBERT, and the NVIDIA Jarvis conversational AI framework. Using NTT Resonant’s proprietary AI algorithms, ConvBERT is able to comprehend the context of multiple statements with high accuracy along with learning from its interactions.
In voice conversations between an AI agent and a human, AI responses are typically dictated by creating rules and using pattern recognition. The demo has a restaurant search and reservation API that identifies the necessary information to make a reservation, including restaurant type and location, time, atmosphere, and more, all while the human and AI converse freely.
The AI will also insert small talk into the conversation, creating the feeling of actually speaking with a human. NTT Resonant provides a type of AI that feels human, accurately interpreting information and user preferences to suggest restaurants without making the user feel like they are interacting with a machine.
