Recho Inc., a developer of a next-generation voice AI platform, has announced that it has filed a patent application for a fundamental technology that will accelerate the practical application of voice AI technology. This patent application technology will go beyond the theoretical possibilities of voice AI and will be the foundation for realizing value creation in real business environments.
The evolution of AI voice-enabled technology and the advantages of Recho
AI voice-enabled technology has evolved significantly, but there are still many limitations. The table below shows a comparison of previous generations of voice AI technology and the unique value that Recho provides.
AI 2.0/LLM voice support has evolved dramatically from conventional technology. Recho‘s platform achieves the fastest response speed in the industry and maintains a natural, human-like conversation rhythm. Specifically, it is possible to handle a variety of corner cases and collect on insurance contracts, which were difficult with conventional AI. In addition, by providing it in PaaS format, adopting companies can easily link with existing CRM and call center systems via APIs while avoiding the need to secure specialized human resources required for AI development and initial investments of hundreds of millions of yen. A major insurance company has achieved concrete results, such as automating more than 80% of customer support within three months of implementation and increasing the payment rate by 1.6 times.
Also Read: Kivotoys Unveils the “AI’mON” with AI and AR Integration
Structural issues in the voice AI market, call centers and telephone response operations
Although the rapid evolution of generative AI has raised expectations for conversational AI, call centers and other telephone response operations still face a structural challenge in that labor costs account for 70-80% of operating costs. This high labor cost rate is not just a matter of cost efficiency, but a fundamental constraint that determines the growth ceiling of the entire business model.
Conventional IVR systems and basic voice recognition technology, known as AI 1.0, have not been able to contribute sufficiently to solving this problem. While many companies are conducting proof of concept (PoC) experiments to introduce voice AI, there are obstacles to moving to the practical stage, such as:
The gap between ideal and reality : Although it works in a limited experimental environment, it cannot support a wide range of use cases in a real business environment.
High operational costs : Even after implementation, significant resources are required for ongoing adjustments and improvements
Complexity of implementation : A wide variety of interaction patterns and exception handling require a high level of expertise to implement at a practical level.
Compared to text-based dialogue, the timing and natural rhythm of responses are particularly important in voice communication, and addressing these points is a major hurdle to practical application.
SOURCE: PRTimes
