Frost & Sullivan, in collaboration with LeadLeo Research, released the '2021 China Dialogue AI Market Report'

A dialogue AI system is a system that understands human intentions in language and performs corresponding tasks or provides answers through speech semantic technologies such as ASR (Automatic Speech Recognition), NLP (Natural Language Processing), and TTS (Text-to-Speech) synthesis. In addition, by combining NLP with machine learning, AI algorithms are continuously upgraded and iterated to help mimic human interactions and achieve efficient intelligent interactive services, such as smart voice assistants and chatbots.

Relying on large-scale deep neural networks, the learning and application efficiency of dialogue AI has been significantly improved, enhancing the experience of dialogue AI systems. With the gradual maturation of NLP's dialogue and question-answering capabilities, knowledge graph technology is accelerating breakthroughs in dialogue AI applications. The product forms of dialogue AI are becoming increasingly diverse, with application scenarios and functions continuously expanding and deepening, including dialogue robots, intelligent voice assistants, in-car voice assistants, and mobile phone voice assistants.

The integration of conversational AI in various industries can be analyzed through four dimensions: data accessibility, technical maturity, scenario penetration, and application importance. Consumer electronics, home appliances, transportation vehicles, e-commerce, and the internet-at-large are the core application scenarios for conversational AI. Applications such as consumer electronics, transportation vehicles, education, government, and healthcare have a high degree of importance, with consumer electronics and transportation vehicles having a high level of scenario penetration.

With the breakthrough in AI voice technology, conversational AI will achieve large-scale penetration in healthcare, education, and government-related application scenarios in the future. Looking at the applications of conversational AI across various fields, there is a high demand for intelligent voice technology in transportation vehicles and consumer electronics. Among the five supporting technologies of conversational AI, speech recognition has the highest demand across industries. The accuracy of speech recognition is crucial as it directly affects the subsequent semantic understanding and intent recognition effects, thereby determining the effectiveness of conversational AI.

• A complete technical ecosystem and efficient development efficiency are the core competitiveness of Chinese dialogue-based AI enterprises.

Dialogic AI enterprises need to continuously invest a large amount of research and development funds to improve technology in intelligent voice semantics and natural language processing, meeting customer needs and expectations. The R&D returns on dialogic AI technology are lagging behind, and the technical system is not yet mature, making it difficult for enterprises to achieve profitability in the short term. Leading enterprises have a more comprehensive range of technologies, gradually transforming from single technology providers to full-stack dialogic AI manufacturers, expanding their industrial chain from top to bottom, and gradually building a complete technological ecosystem.

In terms of industry and application, conversational AI has high demand for industry integration, with personalized customization requirements. For manufacturers with sufficient relevant industry project experience, enhancing large-scale customization capabilities, improving development efficiency for single projects, and reducing operating costs are key to increasing customer volume and average order value. Manufacturers need to provide a user-friendly and easy-to-learn visual operating system, allowing customers to independently build and customize products according to their needs. In industries in the early stages of integration, manufacturers need to optimize service processes, efficiently meet industry demands, and provide standardized products.

• Manufacturers with full-link dialogue AI technology are gradually expanding diversified business models

The profit channels of traditional dialogue AI vendors include providing technical services and hardware fees during the device and technology development process to B2B and B2C customers. In the future, based on full-linkage technical and service capabilities, experience, data accumulation, and user contact and feedback channels, vendors will explore more business models. For example, they can provide integrated software and hardware product solutions for B2B customers, offer more complete solutions and value-added services than algorithms around core needs, help users achieve conversions, and generate profit-sharing. There is also room for exploration in the C2C scenario in terms of content, technology, and promotion, such as providing ecosystem services and personalized customization services.

The Chinese dialogue-based AI market is in a stage of rapid growth. Frost & Sullivan, in collaboration with LeadLeo Research Institute, has divided competing entities into tiers based on their performance in two dimensions: innovation capability and growth capability. The Growth Index measures the competitiveness of competing entities in terms of dialogue-based AI product and service capabilities, market performance, industry scenario application capabilities, etc., while the Innovation Index measures R&D potential, functional innovation, and business scenario penetration capabilities.