The era of AI popularization and application has arrived
Artificial intelligence is one of the most cutting-edge and universal science and technology in this era, and has undergone exploration and innovative applications. At the same time, global IT technology and industry innovation has entered a new stage centered on AI from the previous cloud computing, and business innovation has also entered a new stage of comprehensive digital intelligence business innovation that pays equal attention to digitalization and intelligence. As the world’s leading enterprise digital intelligence software and service provider, Yonyou has conducted research and layout on the popularization and application of AI in advance. A few years ago, Yonyou took the lead in proposing the concepts of “digital intelligence” (digitalization + intelligence) and “digital intelligence enterprise” in the industry, expressing the concept of digital intelligence and intelligence generating new value, as well as digital intelligence business innovation jointly driven by digital and intelligent technology, and a new paradigm of data-driven and intelligent operation enterprises.
Yonyou made a strategic investment 7 years ago, organizing a R team of thousands of people to develop a new generation of digital intelligence for enterprises and public organizations, and achieving a “digital intelligence enterprise” enterprise service product set: Yonyou Business Innovation Platform – Yonyou BIP. On the basis of digitalization, combined with machine learning, natural language processing, computer vision, knowledge graph, IPA (intelligent process automation) and other intelligent technologies, the research and development provides a number of intelligent applications and services, and has been applied in many enterprises.
Wang Wenjing, chairman and CEO of Yonyou Network, said at the press conference: “Generative AI based on large models is the latest development trend in artificial intelligence technology and applications. R training of large models has become a trend in the global industry. In this large model competition boom, Yonyou first chose to be a large model in the enterprise service field that we have always focused on and has advantages, rather than a general language large model. ”
The industry’s first enterprise service model
In order to comply with the wave of AI popularization and application, Yonyou announced the launch of enterprise service large model training at the Yonyou BIP Technology Conference in the spring of 2023.The leading and rich application scenario enterprise software and service product system, the customer base covering dozens of industries, the industry knowledge and experience accumulated by 35 years of serving enterprise customers, and the professional intelligent team and capabilities are the unique advantages of Yonyou’s R and training enterprise service model.
Based on large-scale business application data, combined with enterprise application scenarios and domain experience, Yonyou’s artificial intelligence R team has marked a large amount of enterprise service corpus data, formed rich enterprise service model training materials, and integrated business knowledge and domain experience into enterprise service large models, ensuring the professionalism, practicality and leadership of the large model. Yonyou is also one of the earliest manufacturers in China to apply the Xinchuang framework to develop and deploy standardized AI capabilities.
Fan Guanguan, executive vice president and CTO of Yonyou Network, officially announced at the meeting: Yonyou released the industry’s first enterprise service model YonGPT. YonGPT is an important component of Yonyou Business Innovation Platform – Yonyou BIP, which can adapt to the industry’s mainstream general language large models, and realize enterprise intelligence in four aspects: business operations, human-computer interaction, knowledge generation and application generation by providing general capability services, domain services and industry services.
YonGPT expands the storage memory, adaptation application and scheduling execution capabilities of large models through contextual memory, knowledge/database indexing, prompt engineering, Agent execution, general tool sets, etc., and then combines knowledge and best practices in finance and taxation, human resources, supply chain, R and other fields to expand the professional capabilities of large models, thereby forming a systematic enterprise service model. In addition, YonGPT also optimizes the training efficiency and cost of large enterprise service models, integrates rich development tools and optimization algorithms, and provides stable and effective support for the construction and service of large models through its own data management, large model fine-tuning, large model evaluation and optimization, large model reasoning and plug-in services and other functions. In the future, YonGPT will provide comprehensive capability coverage for Yonyou’s overall product system.
A big model that understands enterprise services better
The application of artificial intelligence based on large models in the field of enterprise services mainly focuses on four directions: intelligent business operations, natural human-computer interaction, intelligent knowledge generation, and semantic application generation. YonGPT, a large enterprise service model, promotes model training and product effect optimization around these four directions, and provides ubiquitous intelligence and swarm intelligence applications that penetrate deep into the front end of customer business and all scenarios.
In terms of intelligent business operations:Through powerful data analysis and prediction capabilities, YonGPT provides in-depth insights into business operations, identifies potential business risks and opportunities, and provides intelligent solutions to improve business decision-making and operational efficiency.
In terms of naturalized human-computer interaction:Through powerful natural language processing technology and understanding capabilities, YonGPT enables enterprise applications and services to engage in natural and smooth dialogue with users, enabling different applications to call, connect, and assemble in a “human” manner, completing work more naturally and efficiently.
In terms of intelligent knowledge generation:By extracting and integrating knowledge from massive data and information, YonGPT generates new and valuable knowledge content, covering industry solutions and knowledge sharing in professional fields, helping enterprises and users fully utilize their knowledge reserves and accumulation, and promote the dissemination and application of knowledge.
In terms of semantic application generation:By understanding user needs, enterprise business, and data characteristics, YonGPT can automatically generate applications with semantic capabilities, comprehensively improving the efficiency of creating personalized application services for enterprises.
At present, YonGPT has innovatively developed dozens of intelligent applications based on the empowerment of large enterprise service models, including enterprise operation insights, intelligent order generation, supplier risk control, dynamic inventory optimization, intelligent talent discovery, intelligent recruitment, intelligent budget analysis, intelligent business travel expense control, code generation, etc.At the same time, YonGPT is aimed at complex industry application scenarios and can also provide more “in-practice” intelligent scenario services by fine-tuning industry models.
Fan Guanjun also focused on four of the intelligent scene services based on YonGPT at the press conference.Among them:
Based on YonGPT’s intelligent analysis of corporate income/profit and tax operations, it can control business conditions in real time, quickly gain insight into problems, accurately predict corporate benefits, and effectively foresee and respond to changes.
YonGPT-based intelligent order generation integrates rich supply chain experience and achieves rapid intelligent order generation through an “interactive innovation” order generation assistant to improve enterprise efficiency.
Based on YonGPT’s intelligent recruitment, it helps companies quickly and accurately identify and locate target talents, discover talents from a massive resume pool, optimize the application experience through AI interaction, and achieve accurate decisions on selecting and employing people.
Based on YonGPT’s intelligent search, it provides a new “immersive” search experience, accelerates the value-based services of enterprise knowledge, and provides insight into user needs, realizes the integration of search and push, and allows knowledge to empower businesses and organizations.
“This release is Yonyou Enterprise Service Model YonGPT 1.0 and the first batch of intelligent scene services based on this large model. In the future, Yonyou will continue to iteratively train YonGPT and continuously enrich various enterprise services based on this large model. Yonyou hopes to continue to serve customers’ digital intelligence construction and operation through the innovative products and services of Yonyou BIP, which includes the Yonyou enterprise service model, so that digital intelligence can be successful in more enterprises and public organizations in China and around the world! ”
