With AI accelerating in the meeting scenario, where is the room for imagination for MAXHUB?

For enterprises, collaboration is the most obvious theme of efficiency, regardless of the presence of AI.

In many inefficient office environments, the biggest bottleneck in collaboration is not the separation between people and data, but the separation between space and people.

Only when AI has good input can it have good output.

Therefore, a large part of the uncertainty in AI's collaborative ability also comes from the lack of spatial perception: the insufficient quality and digitalization of information collection in vision, hearing, touch, etc., seriously hinder the possibility of enterprises and users "using AI right away".

Meetings are one of the very high - frequency scenarios.

As a decision - making link that operates like the "brain" of an enterprise, the complexity of the space, personnel, and data involved in meetings makes it a difficult point in digital transformation. At the same time, meetings are also spatial. The power of AI needs to go beyond the cloud and digital screens and be ubiquitous in the entire physical space and digital system.

"Before AI participates in decision - making, the first thing to do is to enable AI to seamlessly penetrate between people and devices."

Chen Sijiong, the vice - president of MAXHUB, a leading domestic meeting solution provider, told 36Kr that whether it is SaaS such as Teams and Tencent Meeting, or software - hardware integrated solutions like MAXHUB, they are all committed to bridging the gap in spatial collaboration through AI at present.

Recently, MAXHUB released the AI Meeting Solution 2.0, with "spatial intelligence" as the core value, aiming to make AI better assist in meeting scenarios and then meet the needs of enterprises' intelligent transformation.

01 Why can spatial intelligence solve collaboration problems?

In the current era when globalized office is becoming more and more common, remote video conferences have gradually become the mainstream.

However, due to the lack of the concept of "spatial scenario" in face - to - face communication in remote meetings, the limitations of video conferences in information efficiency are also very obvious: transmission delays, restricted speaking due to overlapping voices, and the lack of non - verbal cues such as body language and facial expressions have led to rigid agendas and misunderstandings.

However, even in face - to - face meetings, there are still serious collaboration problems.

A survey by Harvard Business Review in March this year showed that among more than 5,000 knowledge workers, more than 90% of the respondents said that "post - meeting syndrome" had a negative impact on the work process or efficiency: for example, keeping ideas in the mind, repeatedly thinking about fragments of memories, discussing irrelevant topics, lacking clear agenda goals, and lacking follow - up implementation, which led to a sense of frustration.

Therefore, simply having spatial perception ability is not enough. It is also necessary to develop the space towards intelligence, such as summarizing viewpoints, clarifying to - dos, and assisting in decision - making.

MAXHUB has given a practical answer to the problem of the lack of a sense of space in remote meetings through its immersive meeting solution: for example, enhancing information collection in the space through the voiceprint and lip movement of the speaker.

The next step of spatial intelligence is to enhance the input and output capabilities of AI through technological integration on the basis of spatial advantages, narrow the differences in professional abilities, cognitive levels, and language expressions among people, and at the same time run through the entire scenario process to bridge the gap between people, scenarios, and data, so as to achieve human - machine collaboration.

For example, to date, many enterprises still use paper - based or photo - taking methods to retain meeting information, with a relatively low degree of digitalization. The multi - modal understanding ability of AI can silently digitize this content for participants and transmit and retain it to the office scenarios required by the enterprise through Agent - based instructions.

Although the purpose is to "connect people and scenarios", in the technical dimension, spatial intelligence is completely different from tools such as speech - to - text that are already widely used at present.

True spatial intelligence means that AI can participate in and assist in decision - making, and even actively participate in meetings. Chen Sijiong believes that at this stage, what is mainly tested is the manufacturer's accumulation of immersive technology, such as the ability to enhance AI perception.

"In the past, when doing video conferences, we needed to filter out a lot of noise, but the same filtering system may become a negative factor affecting the accuracy for AI. So we spent a lot of effort researching new perception logic."

The advantage of the AI Meeting Solution 2.0 is that in the environment where MAXHUB has achieved cross - device collaboration and high - quality multi - modal input, the rich information dimensions in the spatial scenario can strengthen AI's perception and understanding, and introduce more complete, accurate, and rapid reasoning abilities for applications such as real - time translation, key frame extraction, and meeting summarization, thus helping enterprises create a "ready - to - use" meeting intelligent agent.

02 Technological integration makes spatial collaboration "process - oriented"

Since enterprises started using "large screens", the meeting space has gone through three stages: from the hardware + digital tool stage, to the multi - terminal collaborative stage, and then to the intelligent stage of cognitive enhancement.

Software and hardware are the foundation for intelligent iteration. After the introduction of AI, the meeting space becomes a "thinking" one, which is closer to the nerve center of enterprise decision - making.

Hardware, models, and applications form a "trinity" technical dimension. The challenge of extending human - machine collaboration in the meeting space beyond the space is not only about AI's cognitive ability.

Traditional meeting scenarios are always "disconnected" from office productivity and are regarded as a relatively closed space, mainly because the meeting process is not integrated into the enterprise's productivity process. The collaboration tool chain is extremely fragmented, and different collaboration tools are used before, during, and after the meeting.

Based on this, MAXHUB's AI meeting solution, taking advantage of AI's ability to think and have context memory, runs through meeting reservation, preparation and analysis of meeting materials, communication assistance and record analysis during the meeting, and to - do execution after the meeting.

To achieve consistent performance in the above scenarios, continuous model tuning is very important. MAXHUB uses a large amount of real internal meeting data for training and analysis, and makes targeted optimizations for semantic recognition and understanding in different sub - scenarios such as discussion meetings, training sessions, and interviews, ensuring that environmental perception and information collection are end - to - end, so as to "seamlessly" extend to the office processes before and after the meeting without any learning or process barriers.

Chen Sijiong said that to break the "space" in spatial collaboration and make it flow in the enterprise, a single technological breakthrough is not enough. What is needed is an understanding of the technical architecture and mature technological integration.

MAXHUB understands the intelligent space as four architectural layers, from the back - end to the front - end, namely the decision - making layer, the cognitive layer, the application layer, and the perception layer, corresponding to different technological integrations.

At the perception layer that determines the input, the accuracy and stability of data collection from voice, image, touch, and environmental sensors not only test the quality of the equipment but also its ability to schedule and select the best, which also requires intelligent innovation and iteration.

The same is true for the cognitive layer and the decision - making layer. They not only involve the invocation of multi - modal hybrid models and dynamic knowledge graphs but also need to complete complex recognition and understanding through real - time collaboration protocols and decision - path optimization algorithms.

As the "pioneer" of domestic meeting tablets, MAXHUB has also launched AI + PC products, starting from the meeting scenario to cover the entire process of enterprise intelligent transformation, helping enterprises release their "brainpower" into daily office scenarios and run through the growth of employees' personal abilities.

03 AI applications go deeper into the "new - quality scenarios" of enterprises

The new - quality productivity requires technology to be applied to all industries. For most enterprise customers who only care about "usability" rather than technical principles, creating "new - quality scenarios" is often more effective than creating single - point applications.

As a high - value - density scenario, different from a large number of industrial - end scenarios, although the content, agenda, and professional habits of participants in each meeting are different, the meeting space itself has cross - industry commonalities.

Therefore, what the meeting solution needs is in - depth scenario development rather than industry verticalization.

With the scientific research investment of its parent company, Visionox Technology Co., Ltd., in its "Three Institutes and One Station" and multiple laboratories, as well as the frontier exploration of AI technology, MAXHUB has carried out scenario - based innovation in directions such as high - definition display, human - machine interaction, and artificial intelligence. Supported by the R & D efficiency and existing product capabilities of its parent company, MAXHUB has developed a relatively complete overall solution for the meeting space in China and can focus on developing solutions and serving local deployment instances of customers.

In the past, the biggest difference in meetings across industries lay in different professional corpora, and there were also barriers within different enterprises in the same industry. The MAXHUB AI Meeting Solution 2.0 supports real - time multi - language translation, understanding of industry terms, and can optimize the recognition of professional vocabulary in combination with the enterprise's own corpus, eliminating the pain point of "cross - industry" understanding in meeting scenarios.

This is why, after eight years of dedicated efforts in the meeting field, MAXHUB's customer base already includes more than 80% of the top 500 enterprises in China.

According to QYResearch data, the global sales of the meeting solution market reached $3.923 billion in 2025, with a compound annual growth rate of 8.7% in the next five years, indicating huge market potential. However, the solutions provided by major manufacturers, including Microsoft, Google, and Zoom, still mainly focus on software and SaaS, with product functions concentrated on basic needs such as video conferencing and file sharing.

Undoubtedly, 2024 is the first year of AI applications. With the emergence of innovative models such as a number of virtual meeting platforms, the meeting space is also expected to become one of the earliest scenarios for large - scale implementation of AI in the B - end.

Chen Sijiong also said that MAXHUB is now committed to cultivating the intelligent collaboration habits of enterprise customers from the details of the interaction experience, stimulating the collaborative imagination of AI from the organizational end, rather than "forcing AI to empower".

In terms of software and hardware technologies such as screen displays, sensors, and multi - modal intelligent hardware closely related to the meeting field, the domestic supply chain leads the world, and the integration speed with model application capabilities is very fast. It can also be seen from MAXHUB's product roadmap that the manufacturer has been simplifying the hardware to an all - in - one form to reduce the burden on the enterprise IT department.

As the cost of AI becomes lower and lower, spatial intelligence will also follow the trend of technological inclusiveness and gradually penetrate into more small and micro - enterprises.

What else does spatial intelligence need in the future?

According to MAXHUB's technology cycle, in the coming year, the focus of exploration in spatial intelligence is to standardize multi - modal data and combine applications such as AI Agents to complete more to - dos and generation requirements for enterprises.

In MAXHUB's current meeting AI solution, enterprises can already generate to - do items through spatial intelligence, such as automatically booking the next meeting during a meeting. In addition, in addition to meeting records, AI will also generate key frames for presentation scenarios, etc., to help participants better understand and execute. And this is just the beginning.

Secondly, connecting to the enterprise knowledge base and dynamic knowledge graph can help participants align information gaps and thereby improve real - time collaboration efficiency.

Finally, according to MAXHUB's vision, in the future, AI will also have spatial perception ability like a human being as a digital participant, changing from passive output to active participation in meetings and giving practical decision - making suggestions.