A unifying architecture for easy development, deployment and management of voice-driven mobile applications

摘要

With the advances in voice recognition and synthesis, voice-interactive applications are reaching the mass mobile market. However, creating, deploying or managing context-rich voice-driven applications is complicated due to the scattered and highly diverse nature of specific pieces of software, tools and skills required for a complete system solution. To overcome this, we propose a novel unifying architecture for easy development, deployment and management of voice driven applications which seamlessly integrates mobility and context-awareness, user- and dialog-management, multi-modal interaction and continuous performance measurement and evaluation. Finally, we validate the architecture by stress tests examining possible bottlenecks.