AI agents move from answering questions to carrying out ongoing tasks as enterprise applications expand
Western Securities believes that a new generation of agents can carry out tasks continuously with authorization, opening up potential new opportunities for AI applications. Oracle has launched an agent platform for implementing and upgrading cloud applications, while Chinese companies are also developing products for office work, model access and R&D management.
Original source：财联社First published on Unhidea：Original source date：
Key takeaway
As agents shift from one-off question answering to carrying out tasks continuously, securities firms see commercial opportunities for AI applications, while enterprise product offerings are expanding across office work, model access and R&D management.
How do agents differ from traditional AI assistants?
Western Securities believes that agents are evolving from answering information queries to carrying out tasks continuously, potentially opening up new opportunities for AI applications.
New-generation agents such as Meta Muse use independent cloud environments, long-term memory and connections to tools to handle shopping, travel planning and scheduling with user authorization, while continuing to follow up as external information changes. This frees users from repeatedly communicating or manually intervening at every step.
Bohai Securities believes personal agents have become a strategic priority for leading AI companies. As foundation models improve, personal agents are gradually acquiring the ability to execute real-world tasks, and commercial adoption may accelerate.
Which areas are enterprise applications entering?
Oracle has launched Oracle AI Success Navigator, an agent platform intended to speed up the implementation and modernization of Oracle Fusion Cloud applications.
Weaver Network has launched products including an intelligent foundation platform, intelligent office, intelligent operations and intelligent data capture, covering business scenarios such as sales, procurement, projects and finance. Its XiaoE.AI platform combines large models, RPA and Skill to support multiple agents in initiating, planning and making decisions to handle more complex business interactions.
How are other Chinese companies contributing to the rollout of applications?
Nanxing Machinery has launched the Weiyuan Zhichuang AI Model Cloud Computing Power Platform, which provides aggregated large-model API services. It is also advancing applications in enterprise model access, industry-specific agent development and AI application development.
Digiwin's new-generation PLM system combines text-to-image large-model capabilities with AIGC and AI agents to support enterprise R&D management.
