During the last decade, voice interfaces have become the new way to interact with computers, and over the past several years, we have heard that AI (Deep Learning) has helped speech recognition accuracy reach 95% for specific workloads and applications.
However, usage of voice interfaces is limited and not reliable. Today’s voice assistants help us by bringing content from other applications into their apps, and experience tells us that their reliability is very low; simple tasks are unable to be completed in applications using voice commands, and complex ones are almost impossible. Considering these challenges, it comes as no surprise that almost all applications we use daily are built on touch and type experiences.
The current voice assistants cannot provide a complete and reliable experience for all applications. This is because in the real world, our daily conversations rely heavily on context, whereas these voice assistants do not have any context of the user and the application content.
Humans use not only spoken language but also visual context of the environment to communicate effectively, eg. facial expressions, gestures and content. Using these principles, Alan AI has pioneered the technology to converse with users while leveraging the visual context in any application to provide unprecedented reliability and predictability of conversational voice experiences.
User Context in every application has three dimensions: Visual, Dialog, and Workflow. Visual Context limits the scope of the entity values for AI models, improving the accuracy of interpretation. Dialog and Workflow Contexts give us further insights, to drive better accuracy for predictions made by Alan AI models.
Alan AI is able to provide conversational voice experiences that enable humans to interact with applications of their choice, using their own voices. In other words, for the first time, our innovations make conversational voice experience very reliable and complete, that too, for any application.
The Alan AI Platform brings conversational voice experiences to all applications, using fundamental AI technology for contextual language understanding. With thousands of developers signing up to the self-service platform, we will soon reach the tipping point to make reliable and complete conversational voice experiences ubiquitous, across all applications. Alan AI is poised to be the oxygen required to thrive and excel in this new world.