api.ai Launches Platform Agnostic Voice Interface for Connected Cars
api.ai, a leading conversational interface platform for connected devices, has announced the launch of api.ai for connected cars, a customizable Siri-like voice interface that can be lightly integrated with connected car platforms and services. The api.ai API can be used to incorporate voice-activated intelligent assistant capabilities into navigation, mapping, infotainment, and many other types of in-car applications.
Overall processing flow of a comprehensive speech interface: speech recognition, natural language understanding and conversation management, and fulfillment. – Photo Credit: api.ai
SpeaktoIt, the company behind api.ai, began out five years ago as a voice-activated assistant company. At the time of this writing, Assistant is the highest-rated voice-activated intelligent assistant application available for smartphones and tablets. api.ai is a natural language understanding (NLU), speech recognition, and synthesis platform that was launched this past September. The api.ai platform is based on the Speaktoit natural language processing engine which processes the requests for the Assistant application.
api.ai for connected cars is a platform agnostic voice interface that permits voice-activated intelligent assistant capabilities to be quickly and lightly integrated with in-car systems and applications. The voice interface is very customizable and capable of working with any head unit operating system. The platform includes pre-built voice interface capabilities and permits developers to build custom-made voice-enabled applications.
api.ai has been working with automotive manufacturers as well as OEMs and Tier one suppliers to enable intelligent voice assistant capabilities in connected car platforms and applications. The company has also been talking with several automotive OEMs that provide open platforms and APIs for developers. The idea being to voice-enable open automotive platforms providing APIs that developers can use to build head unit applications.
Ilya Gelfenbeyn, CEO and co-founder of api.ai, told ProgrammableWeb that the majority of feature requests from automotive clients has been regarding navigation and controlling media. api.ai for connected cars is designed so that many of the typical in-car voice interactions are available right out of the box. In fact, Gelfenbeyn told ProgrammableWeb that the company has built their own skill graph and is using machine learning and convolutional neural networks (deep learning) so that the platform can quickly learn based on examples of requests and user interactions.
According to a latest Gartner report, there will be over two hundred fifty million connected vehicles on the road worldwide by 2020. In addition, the number of vehicles tooled with internet connectivity will increase dramatically within the next five years. With so many connected vehicles on roads worldwide, security is a key concern for vehicle manufacturers as well as consumers.
Gelfenbeyn told ProgrammableWeb that security is a concern for the company and that api.ai for connected cars can be implemented as a cloud-based, on premises, or embedded solution (embedded solution depends on the platform). The on premises and embedded implementations help ensure connected car security as they can be used for in-car applications that do not require connectivity or applications where no data is to be sent to the cloud. Gelfenbeyn also said that the company plans on adding extra security features to the API such as OAuth Two.0 and 3-legged OAuth in the future.
“Automakers know their brand and customer better than anyone, so they should be provided with contraptions to design their voice interaction screenplays for in-car practices,” Gelfenbeyn told ProgrammableWeb. “Because of our background in voice assistants that spans over half a decade, we suggest customizable technology that is as sturdy as top solutions presently available on the market. It’s no secret that Automakers stir leisurely, but I’m super excited to see the amazing branded voice interfaces and in-car practices that top brands cook up in-house with our platform.”