Home About us

Why did MiniMax, Dark Side of the Moon, and Zhipu AI choose Feishu?

China Business Strategy 2024/06/25 21:08

Why did China Commercial Towers, MiniMax, Dark Side of the Moon, and Zhipu AI choose Feishu?

Why did China Commercial Towers, MiniMax, Dark Side of the Moon, and Zhipu AI choose Feishu?

Source丨Tiger Sniff APP

Why did China Commercial Towers, MiniMax, Dark Side of the Moon, and Zhipu AI choose Feishu?

In the past year or so, "big model" has been the hottest track in China's technology field.

Especially after entering 2024, the heat of the entire industry can only be described as "crazy": in terms of financing, the boom from 2023 to the present has gradually reached its peak, and the valuations of many leading startups have soared to billions of dollars; At the business level, emerging startups and Internet giants are competing fiercely over the capabilities of the basic model, the price, and the market share determined by the first two.

Even more fierce than the price war is the collision of business routes, with some companies sticking to the business model of to C, focusing on providing products and services directly to the end user. Others are turning to the to B market, providing customized solutions and technical support to other businesses. The remaining enterprises have begun to establish cooperative relationships with Internet giants, seeking broader development space and deeper resource integration.

From technology to application, and then to the fierce collision of commercialization, the entire large model industry is full of gunpowder. But in such an environment, almost all large-scale model startups have reached an agreement on one thing - tacitly choosing Feishu as their collaboration tool without official promotion.

Why did China Commercial Towers, MiniMax, Dark Side of the Moon, and Zhipu AI choose Feishu?

Recently, I went to Guangzhou and saw a customer advertisement for Feishu in the arrival hall of Baiyun Airport, and nine of the hottest AI innovation companies were listed, including the large model "Five Little Tigers": MiniMax, Dark Side of the Moon, Zhipu AI, Zero One Everything, and Baichuan Intelligence.

"Choosing Feishu was a natural process, and I didn't go through too much of a tangled process. It may be because the founding team of large model companies is generally young and more suitable for collaborative tools like Feishu. At the same time, there are too many students from Internet companies in these companies, and they are used to the experience of using Feishu, which has become a mandatory topic for them. A classmate who jumped from the Internet to a large model company said.

Using Feishu seems to be an option that doesn't require much thought for large model companies.

Before analyzing the attractiveness of Feishu to large-scale model startups, it is necessary to understand the actual needs of the latter. If you want to summarize the work status of the large model industry this year in three words, it is "fast, fast, or fast".

Why did China Commercial Towers, MiniMax, Dark Side of the Moon, and Zhipu AI choose Feishu?▲ The domestic large model has been surpassed in a variety of capabilities.

Picture from: FlagEval

In order not to fall behind in the track of large models, thousands or even tens of thousands of GPUs are often used for training, and updates are made faster and faster. In terms of basic large models, it is not only necessary to catch up with and surpass foreign advanced models in terms of performance, but also to compete fiercely with other domestic competitors in terms of application landing. Many companies have also entered the commercialization stage ahead of schedule, trying to be the first to explore the market and find customers in the to C and to B markets to realize business value.

At the same time, the arduous task of promoting "scientific research, application exploration and commercialization" has allowed the large-scale model startups that have been established for only 1~3 years to enter a period of extremely high-speed expansion, with larger teams, more projects, and more complex commercialization attempts, which directly caused a surge in the demand for collaboration.

Let's talk about the challenge of "people" first, there are many "super individuals" (employees with outstanding professional ability) in the R&D team of large-scale model startups. Take the dark side of the moon, one of the "Five Little Tigers" of the large model, as an example, its early core team is all post-90s and from Tsinghua University, and it has been rated as one of the five candidates most likely to become "China's version of OpenAI" just three months after its establishment. The dependence of the entire large-scale model industry on "super individuals" has posed a new challenge to enterprises from the initial recruitment to the later management incentives.

In the face of these "super-individual" employees, it is impossible to apply the single dimension of talent evaluation based on results. Second, for startups that are highly young and full of "super individuals", the traditional methods of attendance and solidification processes are also out of place. Large-scale startups urgently need a more flexible and innovative way to coordinate and improve the efficiency of "people".

From the perspective of management, the large model industry is still in the stage of rapid development and needs continuous exploration and practice. This process requires the joint efforts of all functional departments of the company, "crossing the river by feeling the stones", and more important than implementing the instructions of the boss is to give full play to the "creativity" and "initiative" of each employee, quickly respond to changes in the market and technology, and promote the development of the project.

If the production and research of large models is already very difficult, then the commercialization challenges that large model startups need to face today can be described as "more difficult".

How to continuously improve the level of capability at the level of product research and development, how to plan the new functions that individual users really need, how to work with enterprises to explore the application potential of large models, and how to adjust their own development strategies based on the real feedback from the market; The emergence of this series of new problems has made the overall challenges faced by large-scale model startups more and more "complex".

Taking Zhipu as an example, it was officially announced at the beginning of 2024 that it had 1000+ large-scale models for large-scale application and carried out in-depth "co-creation" with 200+ enterprises. This process of "co-creation" is actually the process of "excavating" the needs and application prospects of large models in all walks of life together with customers, which has generated a large number of communication and collaboration needs between Zhipu and customers, sales and development, and between the front line and management.

Obviously, at the current critical time node when the large-scale model technology is gradually maturing and more and more attention is paid to the application of the landing, the large-scale model startups are chasing the whole chain of innovation from theory, method, technology, product to market, coupled with the uncertainty brought about by the rapid development of the industry itself, the huge challenge of pushing a new technology to all walks of life has formed a strong sense of urgency that lingers in the hearts of large-scale model startups.

Why Feishu? The author's speculation from the characteristics of the large model company probably boils down to three points:

1. Rapid iteration of tools drives the iteration of the organization:

In the process of rapid expansion, large model companies urgently need mature methodologies and tools to improve efficiency and even iterate the internal operation mechanism. It just so happens that Feishu has accumulated the methodologies of thousands of advanced enterprises in the past few years, and has brought positive changes to the organization in the rapid iteration of tools.

2. context not control:

Building an agile organization is the goal of an emerging company like the Big Model out of consideration for a rapidly changing business fit. The all-in-one model adds to the rapid flow of information and achieves efficient organizational synergy. The so-called context not control, today's large model companies are goal-driven, more like distributed organizations, bringing out the potential of each individual.

3. High flexibility and openness of products:

Feishu's multi-dimensional forms and open platform capabilities have strong openness and creative space. This feature just matches the characteristics of large model companies. Such companies are highly concentrated in production and research, are good at using tools, and have built many applications by themselves through Feishu's open capabilities. It is said that the dark side of the moon has used the Feishu integration platform to build dozens of enterprises' own applications, which are integrated with other functions on Feishu to jointly improve efficiency.

For example, ordinary IM and collaboration tools can realize the flow of information within the scope of their respective products, while Feishu's system can realize the connection between multiple different functional products, specific projects and approvals, and processes and knowledge precipitation.

This ability to speed up information transfer is especially effective in complex scenes.

For example, traditional CRM software is good at analyzing customers' historical transaction data, but it does not have powerful communication and office functions, and they may become information silos in the long-term use of enterprises, resulting in inconvenience and process inefficiency. Feishu has enhanced the ability to import and interact with sales data, making data management more efficient, and helping sales staff make timely decisions and effective follow-up with more macro and clearer data presentation and insight capabilities, improving efficiency and performance.

Why did China Commercial Towers, MiniMax, Dark Side of the Moon, and Zhipu AI choose Feishu?

In addition to accelerating various collaboration processes within the company, Feishu's "information-centric architecture" feature has also changed the dilemma of business experience following individuals and not being able to use individual wisdom to drive group development. As we mentioned above, the information about the entire project is aggregated into a knowledge base and connected by a page tree structure that is much clearer than a folder system. Even new employees who have not worked on this project will be able to learn from historical documents afterwards and summarize the implementation process.

By improving the overall efficiency of knowledge dissemination within the enterprise and transforming information and experience into a systematic knowledge base, large-scale model startups can efficiently circulate and utilize this knowledge. This systematic process not only promotes the collision of ideas and innovation among employees, but also feeds back these innovative results to the enterprise itself, which has become an important driving force for the development of the enterprise.

If the choice of Feishu at the beginning is a judgment of "all based on feeling", then considering the key issues such as going overseas and safety, it seems that Feishu has become a choice that large model companies have to make.

On the one hand, in the long run, almost all large-scale model companies have the need to go overseas, and considering the various cross-country and regional synergies and strict cross-border compliance requirements, it seems that only Feishu, which serves many enterprises to achieve a global layout, can meet them. On the other hand, the extreme requirements of security also preclude some collaborative software options for large model companies. At present, Feishu has done an extremely thorough job in terminal security, data encryption label protection, data leakage prevention and other parts. One of the reasons why many large-scale startups like to use Feishu is that it is very "meticulous" in its management authority for various types of information, and can even limit the scenarios of replication and copy creation on demand, and the right to share information externally can be different according to the scenario.

In fact, if you broaden your horizons, the advanced companies that choose to use Feishu are far more than just startups that focus on large models. From autonomous driving to embodied intelligence, to the upstream and downstream of AI innovation companies, Feishu seems to have always been the common choice of "advanced industries" at the forefront of this society.

Feishu's advanced collaboration capabilities are not only appealing to companies, but also to individual users who represent the most advanced productivity. Just like those "old users" of Feishu in large-scale model startups, they are expanding Feishu's influence and territory outward like a "torch relay".

WaytoAGI (The Road to Artificial General Intelligence), the most influential open-source AI knowledge base project in China, built on the Feishu knowledge base, is the best example.

▲WaytoAGI diagram

Since its establishment in April 2023, WaytoAGI (The Road to Artificial General Intelligence) has built an extremely large knowledge base system, covering various technical introductions, AI industry news analysis, and AI application practice, which has received more than 1.5 million views, triggered tens of thousands of exchanges between users, and enabled at least hundreds of thousands of Feishu users to have a deeper understanding of AI and make their own attempts to apply AI. The fact that a group of people with lofty ideals in the network can achieve such results not only proves the excellence of Feishu's entire collaboration system, but also proves that "people who understand large models are likely to use Feishu." ”

On the premise that it already has a large number of "fans", Feishu has not stopped its own evolution.

While serving the customers of large-scale startups, Feishu has also begun to accelerate the introduction of AI capabilities into its products. As early as the end of last year, Feishu has embedded the "Feishu Smart Partner" in its entire set of products, which can work with users in business scenarios such as content creation, content summarization, data analysis, scenario construction, and system construction. Enterprises can even choose suitable underlying large models according to business needs, such as Baichuan Intelligence, MiniMax, and Zhipu AI.

Take the daily work summary, which is the most common work of employees, as an example, which can be directly generated by the intelligent partner. Some PDF files with complex content and a large number of words can be directly sent to the smart partner to extract and summarize the key content in the form of questions and answers. You can even describe your needs and uses in words, so that your smart partner can build the required multidimensional spreadsheet system.

From advanced customers, to advanced users, to all AI-related capability upgrades, perhaps we can say that in China, everything related to AI is becoming more and more deeply related to Feishu.

This article is from Xinzhi self-media and does not represent the views and positions of Business Xinzhi.If there is any suspicion of infringement, please contact the administrator of the Business News Platform.Contact: system@shangyexinzhi.com