Agora, Inc., a provider of real-time engagement solutions, has announced the launch of its ConvoAI Device Kit, targeting the development of AI-powered robotics and interactive toys. This toolkit is designed to enable developers and manufacturers to integrate voice-driven interactions into smart devices.
The ConvoAI Device Kit is the result of a partnership with chip-maker Beken. It combines Beken’s AI chip modules with Agora’s Conversational AI technology. Key features include ultra-low latency voice interactions, intelligent dialogue processing, and real-time communication capabilities. The stated goal is to make AI-driven devices, including toys and robots, more interactive and responsive.
Key partnerships and applications
One of the first applications of this technology is in Robopoet’s new AI companion robot, Fuzzoo, which was showcased at the recent Mobile World Congress (MWC). Fuzzoo is designed as an emotional companion robot, using Agora’s technology to provide real-time listening, sensing, and response capabilities. Robopoet’s Multimodal Emotion Model (MEM) utilizes the ConvoAI Device Kit for personalized emotional support.
AuditBoard unveils AI-powered features for internal audit
Agora highlights several potential use cases for its AI device solution:
- Educational AI toys: Supporting STEM learning, coding, and problem-solving.
- Companionship devices: Designed for emotional interaction and support.
- Interactive play toys: Featuring conversational AI with voice and touch interactivity.
- Smart home devices: Incorporating advanced voice AI agents.
- AI-powered wearables: Utilizing smart AI voice interaction.
A key aspect of Agora’s approach is the direct integration of AI into devices through its partnership with Beken on edge chips. This on-device conversational AI is presented as a solution to common challenges like background noise interference, latency issues, and the limitations of rigid AI models. Agora’s technology addresses these issues with features such as voice activity detection (VAD), real-time speech synthesis, and intelligent interruption handling. The overall aim is to make it more detailed and complete.
Featured image credit: Agora
