Shulou(Shulou.com)11/24 Report--

Large models shine brilliantly in the field of natural language processing, and documents are one of the most important carriers of natural language, so when large models become a new generation of "infrastructure", intelligent documents become the first track to usher in change. it is also the fastest to launch a market-oriented product category.

A few days ago, Tongxin Software officially released UOS AI, announcing that the community version is connected to the big model, and has enabled three self-developed applications. Immediately, the domestic well-known online collaborative office software brand wrote freely and announced that its products had access to UOS AI online, becoming the first third-party application in the world to access UOS public AI underlying services.

Start with three major functions

Building Intelligent document Productivity

Generally speaking, document types can be divided into streaming documents (suffixes such as doc, docx, wps) and layout documents (suffixes such as pdf and ofd). Streaming documents have the characteristics of free typesetting, and content generation can be completed by using large models of natural language, so office software have chosen to take the lead in testing the water in their online document products.

Through UOS AI, open write online realizes three kinds of content generation capabilities:

First: the outline of the document from 0 to 1.

Everything is difficult at the beginning. If you want to improve office efficiency, the key is to take the first step. Create a new document in the popular online, call UOS AI, select the document outline, enter the title, and UOS AI can generate a complete article outline.

Second: intelligent creation of documents.

With the help of UOS AI, AI continuation, translation, language error correction, text retouching, content expansion and summary can be realized in the editing process. Until you get a satisfactory result, click finish and you can insert it into the text of the article.

Third: AIGC multi-round dialogue

In the document built-in real-time dialogue AI tool, can carry out multiple rounds of dialogue. Users can ask questions at any time in the document, which changes the interactive interface and way of traditional online documents, which is equivalent to the real-time assistance of an intelligent document assistant to make the creation more focused.

"We hope that before users start any creation, there will be a brainstorming session with AI to stimulate creative inspiration and get into work more quickly." Shi Zhoubo, general manager of Changwen online, said that when going deep into industry areas and professional scenarios, UOS AI can also give users more goal-oriented assistance.

For example

In the ToG document scene, UOS AI can help generate a variety of templates such as meeting notices, administrative documents, personnel changes, and can be adjusted to a more formal and standard language style.

In the industry, UOS AI can assist in the generation of reports in insurance, banking, securities, investment and other sub-areas.

In the general office field, UOS AI can assist product analysis, competitive product analysis, report search, report generation and so on.

Three stages

Realize AI deep enabling

For users, AI enabling creation brings a new office experience, and more application scenarios need to be explored. As for product research and development, there are more technical difficulties to be broken through in the integration of AI and documentation.

Shi Zhoubo believes that content generation is only the first step in the long march.

The first is data security:

In the internal test phase, the online UOS AI calls the public cloud API, which is not suitable for data storage. At present, the well-written AI R & D team is working with Uni-Credit software and large model manufacturers to privatize and deploy the access and scheduling framework of large model, industry exclusive model and small model, in order to better meet the office demands of the party and government and key industries.

The second is the product perfection of smart documents:

On the one hand, improve the processing ability of intelligent documents, invest in research and development in the analysis and restoration of layout documents, information analysis and understanding of rich text documents; on the other hand, open up the document-based business flow, build AI office ecology, and work with partners to promote process AI reshaping, further liberating manpower, reducing costs and increasing efficiency for the digital transformation of enterprises.

The third is the construction of knowledge base:

When the large model goes deep into the business scene, in the face of more diverse data types and more complex data distribution, the construction of the knowledge base will help AI to extract knowledge more accurately and form reuse more effectively, and the training effect of the exclusive model in the vertical domain will be better.

Zhang Lei, senior deputy general manager of Tongxin Software and CTO, said that freely writing online access to UOS AI is a practical test of the underlying public capabilities of the operating system AI, and also provides prompts and other feedback for UOS AI. At present, UOS AI has achieved GPT, iFLYTEK, Baidu, 360, Zhiyuan and other multi-model access capabilities, enabling the application of AI. With the gradual open source of UOS AI, it is hoped that more third-party applications can directly call the large model on our operating system.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.