Microsoft China Strategic Incubator Zhang Siyuan: Advancing Technology and Applications, Pioneering a New Global Entrepreneurship Paradigm

Recently, the 2026 Chongli Forum was held at Taifeng Town in Chongli District, Zhangjiakou City. Zhang Siyuan, General Manager of Microsoft China Strategic Incubator, was invited to attend the exclusive roundtable session, where he shared insights on how incubators judge entrepreneurial teams and directions in the AI era, providing forward-looking thoughts on technological evolution, entrepreneurial pathways, and investment opportunities for entrepreneurs and investors both on-site and online.

The Chongli Forum is characterized by its integration of natural scenery and cutting-edge thinking, focusing on key topics such as artificial intelligence, digital economy, and green low-carbon development. It has become an important annual cross-industry dialogue platform for China’s tech and industrial sectors. In 2026, the forum’s theme is “Standing at the New,” gathering nearly a hundred heavyweight guests from the tech, industry, and capital sectors to discuss the new order of the intelligent era.

Centered around the topic “Has AI Investment Entered the ‘Post-Consensus’ Stage?” Wang Lu, Partner at Guochen Venture Capital, moderated the roundtable discussion. Participants included Zhang Siyuan, General Manager of Microsoft China Strategic Incubator; Jiang Chun, Managing Partner at Puhua Capital; Zhang Qian, Founding Partner at Tianji Capital; and Wu Bingjian, Partner at Xin Capital. They engaged in in-depth exchanges on core issues such as the evolution of large model technology, differentiated paths for startups, commercialization, and investment criteria.

Below are excerpts of Zhang Siyuan’s viewpoints from the roundtable discussion:

01

2026

What kind of entrepreneurial teams does the incubator focus on?

We mainly focus on two types of teams: Chinese entrepreneurial teams with a global perspective from Day One, and entrepreneurs with a solid research background (Researcher-led).

After validating TPF (Technology Product Fit) and PMF (Product Market Fit), startups often enter the market by addressing specific user pain points or scenarios. However, in an increasingly competitive environment, relying solely on a single market or regional advantage may limit future growth. Therefore, teams that are founded with a global market orientation have greater potential and imagination for scaling.

Meanwhile, AGI has become a long-term goal widely recognized in the industry, but there are still multiple explorations on how to achieve it. True breakthroughs may come from innovations in underlying methods and architectural paradigms—such as the Transformer architecture supporting large-scale parallel training, which has driven leaps in model capabilities. Based on this judgment, we are proactively engaging with global research institutions and research-oriented startups to seek paradigm shifts.

02

What entrepreneurial directions does the incubator focus on?

First, AI application directions. Chinese entrepreneurs have obvious advantages in AI applications, stemming from their experience in high-intensity competition during the mobile internet era, which has cultivated keen user insights, solid product refinement capabilities, and mature, efficient engineering and user growth skills. This characteristic is also reflected in the incubator’s portfolio companies. Currently, about half of the startups focus on AI applications. Many of these teams do not have long-term overseas backgrounds but have built global teams and iterated products rapidly, receiving enthusiastic responses in overseas markets. For example, Lovart attracted over 800,000 global users during its beta phase, and after official launch, its ARR soared, demonstrating the competitiveness of Chinese entrepreneurs in AI application innovation.

Of course, our focus on applications is also supported by Microsoft’s solid accumulation in infrastructure, security systems, and key protocols and platforms, which continue to serve as the foundation and enabler within the overall ecosystem. With this full-stack capability, we can provide systematic support to application startups in technical assistance, platform integration, and global market expansion, helping them reach international markets more efficiently.

Second, consumer goods companies leveraging China’s supply chain advantages and targeting global markets. These companies rely on the flexible supply chain systems of the Yangtze River Delta and Pearl River Delta regions, as well as mature smart hardware ecosystems, giving them outstanding advantages in product design, cost control, and delivery efficiency. From apparel to smart hardware, many Chinese teams have achieved commercial success in global markets. As AI capabilities continue to integrate into hardware products, domestic entrepreneurs are creating differentiated experiences through integrated soft and hardware design. For example, Plaud AI combines large model capabilities with conference recording hardware, achieving over a million units shipped worldwide. This year, we will also see more AI companion hardware products going global.

03

In 2025, AI applications will emerge in abundance

Have you already identified some definitive opportunities?

Indeed, 2025 will see an explosion of AI applications, and the perspectives of investors and entrepreneurs on AI application “wrapping” have also shifted. Through conversations with many incubated companies, I found that the key behind “wrapping” still lies in whether the startup team truly understands users, accurately grasps core needs, and continuously deepens product functions, interactions, and value delivery. The goal is for users to remember the product itself, not just the underlying model capabilities. Once the product’s value is validated by the market, being acquired by a giant is a good exit route, and it also signifies that the startup team has established a clear and advantageous market position amid competition.

This year, AI applications will continue to deepen across multiple directions. Rather than focusing on opportunities within specific vertical industries or scenarios, I prefer to approach this from the user perspective. Currently, a relatively clear path is to prioritize serving prosumers—productive individual users. These users have a certain level of technical understanding, are quick to adopt, tolerant of errors, willing to explore new products, and willing to share or subscribe when they see real value.

In the early stages, focusing on prosumers not only helps validate product value and user needs but also generates revenue and supports subsequent fundraising. Once the product gains traction among this group, startups can leverage their own product features, team strengths, and input-output efficiency to find new growth curves: some teams may move into more complex but higher-value enterprise markets (To B), while those skilled in consumer applications can further lower usage barriers and increase engagement to expand into broader C-end markets.