Original · Unique Research · 2026-03-25 · Shanghai
Editor's note: This is a historical summit recap, although the source opens with “EVENT PREVIEW.” Event scale, policy descriptions, business outcomes and medical figures below are claims in the original report or attributed speakers' remarks, not independently verified results. The OPC policy was a draft for comment at the time; the source does not specify the currency of its industrial-fund figure. The cited 96.5% dental assisted-diagnosis accuracy has no evaluation method or sample details in this report and should not be read as established clinical performance. Some organizations and people are rendered by romanization or descriptive translation pending confirmation of their official English names. This recap calls Procen “Beijing,” whereas the separate March 23 interview calls it “Xi'an”; neither location has been silently substituted. Security and AI-governance positions are reported views, not recommendations to relax safeguards.
EVENT PREVIEW
2026 Unique Awards · Hangzhou AI Week |
The Unique Capital AI Entrepreneurship and Investment Summit Concludes
When AI Truly Begins Entering Industries, Organizations, Workflows and Jobs
On March 25, the Unique Capital AI Entrepreneurship and Investment Summit, part of 2026 Unique Awards · Hangzhou & Hangzhou AI Week, took place in Hangzhou's Binjiang district. It was organized by Unique Research and co-organized by the Hangzhou Association for the Promotion of Science, Technology and Cultural Industries.
The gathering was more than a collection of people interested in AI; it resembled a rapidly forming network of industry relationships. With 3,000+ registrations, 1,000+ AI companies, 500+ founders and 500+ investment institutions, Hangzhou that day offered a concentrated view of Chinese AI entrepreneurship. Investors recalibrated their understanding of industry, founders shared fresh experiences of deploying applications, and dense discussions covered cross-border business, healthcare, office work and enterprise services, alongside new judgments about future entrepreneurial models.
From the morning keynotes through a full day of trends roundtables and the afternoon's parallel Unique DemoDay, the summit had moved beyond grand questions such as whether AI would change the world. It addressed something more practical:
As AI truly begins entering industries, organizations, workflows and jobs,
what exactly should founders, investors, companies and platforms
do next?
Starting with Highly Leveraged Individuals, Hangzhou Offers a New Entry Point into AI Entrepreneurship
The summit opened with a keynote by Shui Xuefei, Vice President of DBAPPSecurity, titled "The UniWis OPC Community: A Paradigm for Highly Leveraged Individual Entrepreneurship in the AI Era."
The presentation was not merely a community launch. It sought to define a new entrepreneurial unit closer to today's AI reality. Shui introduced the UniWis OPC community, jointly initiated by DBAPPSecurity and Unique Research. Located in DBAPPSecurity's Phase II Science and Technology Innovation Building, the community launched as Binjiang district introduced a draft-for-comment policy supporting OPC entrepreneurship, including rent support, computing vouchers and R&D subsidies. For AI entrepreneurs moving from isolated capabilities to organized delivery, this is more than workspace: it is an attempt to package policy, technology, an ecosystem and ongoing support into entrepreneurial infrastructure.
Shui argued that entrepreneurship in the AI era is no longer confined to one-person companies. It extends to highly leveraged individuals and small teams better suited to AI's multiplying effect. In his view, founders now need the capabilities of a product manager, a systems architect and an intelligent orchestrator: understanding users, designing systems and mobilizing AI to complete tasks. The talk also outlined a clear path: Future entrepreneurship is not simply about what I can do, but what I can coordinate, integrate and amplify.
Next, Yan Luojia delivered a welcome address, "From Policy to Industry: Exploring AI Entrepreneurship Ecosystems and Capital Collaboration." She is President of the Hangzhou Association for the Promotion of Science, Technology and Cultural Industries, Executive Chair of the Zhejiang Entrepreneurs Association's Digital Assets and Trading Committee, Chair of Zhejiang Anchuang Zhilian Technology Co., Ltd., and a Director of Zhibaixing Cell Bank (Zhejiang) Co., Ltd.
Yan focused on Hangzhou. She said the city's appeal to AI founders is not only market enthusiasm, but a more complete support system taking shape: policy, capital, talent, industrial chains and scientific resources are increasingly converging around AI. Industrial funds on a 100-billion scale, the science and technology bureau's Runmiao program, universities such as Zhejiang University and Westlake University, and local research resources all provide fertile ground for AI development. She stressed that the industry needs not detached capital, but patient capital that understands technology cycles and accompanies projects as they grow. Capital should help build industries, not merely watch them.
Warren Li, Head of Venture Ecosystem Business at Google Cloud, then delivered the keynote "What to Pursue—and What Not to Pursue—in Venture Markets, Viewed Through the Google AI Roadmap."
Starting with Google's AI roadmap, Warren Li brought the discussion back from fashionable concepts to structural questions that genuinely affect entrepreneurial success. He reviewed Google's work in TPU, Gemini, multimodal models and enterprise AI, then argued that founders should not chase the giants' most crowded battlefields. They should find poorly served niches with extremely low industry efficiency, outside the coverage of major-company ecosystems. Rather than compete for attention in hot sectors, rebuild an overlooked industry from first principles. Warren Li also emphasized trust, security and customization in the ToB market, and the significant opportunities still available to Chinese founders in enterprise services. The morning quickly developed a shared conclusion: AI opportunities are not disappearing; they are moving from buzzwords toward structural gaps.
As Investment Logic Is Rewritten, Industry and Capital Are Realigning
The first trends roundtable, "Industry and Capital: Reconstructing Investment Logic and Discovering Value in the AI Era," was moderated by Wang Chunfeng, Secretary-General for Incubators at the Hangzhou Science and Technology Innovation and Entrepreneurship Association. Guests included Wu Wei, Founder of Unique Capital; Ni Min, Executive President of Zheshang Venture Capital; Wang Aiwu, Founder of Longqi Investment; and Yuan Zhiyong, Partner at Saizhi Bole.
A strong signal from the discussion was that AI investment is moving beyond who can tell a story toward who can truly become embedded in industry. Starting with Zhejiang manufacturing, Wang Aiwu discussed the trend illustrated by Geely and other cases: Hangzhou's software capabilities and Zhejiang's manufacturing base are forming a new connection, making AI a question of industrial upgrading as well as software. Yuan Zhiyong focused on how company value should be reassessed once AI becomes a digital employee. He argued that technology investment's fundamentals remain: sector, team and market size still matter, but investors must pay closer attention to whether AI genuinely creates efficiency in vertical settings such as industry, healthcare and law.
Ni Min's view was direct: AI investment is shifting from concepts to infrastructure and from technical showmanship to practical use. Computing, storage and other AI infrastructure offer greater certainty, while AI's lowering of entrepreneurial barriers may make individuals and small teams more important sources of innovation. From a perspective closer to founders' daily work, Wu Wei identified three entry points worth watching: data, the medium or device carrying the experience, and interaction. He also acknowledged AI's security and ethical risks, while emphasizing that founders must recognize speed as a primary productive force: Half a step too slow may mean missing the window.
The second roundtable, "Differences in Chinese and US AI Industry Narratives and Their Implications for Capital," was moderated by Tina of Unicorn Interview Room. Guests included Wayne, Managing Partner at Argo Venture Partners; Piruze Sabuncu of Square Peg; Hu Bin, Founding Partner at Yingce Capital; Cao Wei, Partner at Lanchi Ventures; and Shen Dongliang, Founding Partner at Yuanshu Venture Capital.
The value of the discussion was that it did not speak vaguely about differences between China and the US. It situated them within specific industry structures, technical paths and investment judgments. Wayne, Piruze Sabuncu, Hu Bin, Cao Wei and Shen Dongliang examined AI's changes from perspectives spanning North America, Asia and Australia, early-stage domestic investment and international expansion. They generally believed frameworks such as the lobsters are changing production logic in the AI application layer: Many future applications will serve machines rather than people. Competition will become more intense, and real barriers will increasingly come from industry know-how, proprietary data and reusable skill nodes.
The guests also drew a relatively clear conclusion: China and the US are not simply ahead or behind one another; each has strengths. Overseas ecosystems are more open in model-architecture and training-paradigm innovation, while China is more agile in hardware manufacturing, engineering deployment and embracing open source. The window for Chinese AI companies to expand abroad has opened, but they cannot merely transplant Chinese solutions. They need to consider global users' problems from the beginning. In business and organizational design, separating Chinese and US operations, maintaining compliance separation and managing geopolitical risks are also becoming unavoidable questions for companies seeking long-term success.
The morning's third roundtable, "Differences in AI Entrepreneurship Opportunities and Cross-Generational Collaborative Innovation," was moderated by Li Jinxiang, Founder and CEO of Xiniu Data. Guests were Wu Jiabing, General Manager of Strategic Investment at Wondershare; Zhao Peizhou, Partner at Xiaomiao Langcheng; Lu Hongyu, Director and Senior Partner at Detong Capital; Ren Bobing, General Manager of the Frontier Technology Fund at Sinovation Ventures; and Bu Liangyuan of Shidai Bole's Listed-Company Investment Department.
One highlight was how investors from different generations and institution types viewed today's AI founders. Ren Bobing noted that, compared with the AI 1.0 period, founders now have more mature infrastructure and clearer technical starting points, yet the industry remains full of unresolved questions, much like the early internet. Zhao Peizhou described two promising founder types: university professors with deep research and systematic capabilities, and younger founders able to judge quickly and seize windows of opportunity. Lu Hongyu reminded the audience that investment's commercial fundamentals have not changed. However new the concept, the questions remain whether the market accepts the product and whether users will buy again.
A consensus gradually emerged around advancing hardware and software together. On hardware, Zhao Peizhou and Bu Liangyuan favored AI hardware for home and everyday-life scenarios, wearables and embodied intelligence. On software, Ren Bobing, Wu Jiabing, Lu Hongyu and others saw opportunities in agents, world models/event models, lobster-like intelligent-assistant ecosystems, and content sectors such as AI animated series and AI music. The summit sent a clear signal: AI entrepreneurship opportunities in 2026 will not belong to one technical trend alone, but more likely to those who genuinely embed technology in specific scenarios.
At the morning's final highlight, Unique Research and Xiniu Data officially released the "2026 AI Investors 50" list. It was not only a ranking announcement, but also a concentrated presentation of judgment in Chinese AI investment over the preceding year.
In the Afternoon, AI Became a Question of Transforming Specific Industries, Not Just Technology
The afternoon discussions clearly shifted from broad judgments to vertical deployment.
Zhang Xuguang, Chair of Hangzhou Juexingdao Artificial Intelligence, moderated "AI and Industry: Paths and Practices for Intelligent Transformation in Traditional Sectors." Six very different practitioners participated: Yang Wenjun, Chair of Zhejiang Kangbaiyu Biotechnology; Hua Shaobing, Chair of Hangzhou Detong Biotechnology; Guo Yuchen, Chair of Zhejiang Xialinghui Intelligent Health and Elder Care; Heiyu, Founder of Dayou Space; Li Qinfeng, Founder of Yuanji Digital Intelligence Technology; and Xiao Yineng, Director of the Intelligent Education Laboratory at Peking University's Advanced Institute of Information Technology.
The most interesting aspect was that the speakers did not understand AI as one uniform thing. Yang Wenjun discussed AI in biopharmaceutical R&D, making experiments and research processes more precise and efficient. Hua Shaobing described AI working with pathology testing and biomedicine in women's health diagnosis. Guo Yuchen placed AI in health, elder care and the silver economy, emphasizing robots' emotional companionship and spiritual value as well as functionality. Heiyu focused on new relationships between virtual environments, interactive experiences and e-commerce conversion. Li Qinfeng described how consumer products such as smart bead bracelets turn agents into perceptible features. From a research perspective, Xiao Yineng cautioned that embracing open-source models and new capabilities must also include attention to security vulnerabilities and technical uncertainty.
If the morning was largely about what capital looks for, this roundtable answered another question: Why do traditional industries need AI, and in what form will it enter them? The conclusion was concrete. AI is not one template copied across sectors. Different industry characteristics, risk tolerance, data conditions and organizational structures require different product forms and deployment logic.
The next discussion, “Medicine and Health: Innovative Applications of AI Diagnostic Technology and Intelligent Health Assistants,” was moderated by Jeffrey Wu Shenliang, VP at Unique Capital. The panelists were Pan Shouxiang, CFO of Quanzhen Medicine; Qiao Minghui, founder and CEO of Yuemi Technology; Sun Junwei, CMO of Hengfang Health; and Wang Pu, co-founder of Beijing Procen Intelligent Medical.
The discussion was restrained, which made it all the more substantial. All four panelists represented companies already implementing products in real medical settings: hospital workflows at top-tier tertiary hospitals, pharmaceutical marketing and patient management, agents serving doctors and patients, and specialized disease AI and imaging-assisted diagnosis. Their underlying consensus was clear: medical AI must not be mythologized; it must return to real settings. Pan Shouxiang discussed breaking doctors’ daily workflows into sufficiently fine-grained steps so AI could assist rather than replace them. Qiao Minghui emphasized that AI breaks the old triangle in which personalization, scale and low cost were difficult to achieve together. Sun Junwei argued that the real threshold for medical AI is not merely being “useful” but being “easy to use.” Wang Pu cited a 96.5% assisted-diagnosis accuracy rate for dental AI to illustrate the continuing value of specialized models in narrowly defined disease settings.
More importantly, the discussion offered a clear-headed assessment: faced with the rapid iteration of general-purpose large models, the medical industry will neither become blindly optimistic nor stop moving forward. The seriousness of medical settings means AI must be deeply integrated with clinical workflows, data compliance and safety systems. Rather than a race over parameter counts, what truly determines success is deep specialization in vertical settings, engineering capability and compliance capability. This also turns “AI+ medicine” from an imagined grand narrative into a real industry whose development can be broken down, verified and advanced step by step.
As AI Takes Over Knowledge, Marketing and Office Work, What Companies Really Need to Rebuild Is Their North Star Metric
The roundtable “Learning and Knowledge: Knowledge Management and Productivity with AI as a ‘Second Brain’” was moderated by Wu Wei, founder and CEO of Unique Research. Panelists were Wang Baochen, founder and CEO of Chuangkit; Chen Ming, founder of NomiLaw Nuomibao; Yu Chunyan, co-founder of Xuntu Technology; Wu Bin, CEO of Jirui Technology; and Ning Liaoyuan, CTO of TTC.
The discussion felt like a collective retrospective on rewriting companies’ “North Star metrics.” The panelists came from design tools, legal and tax compliance, financial investment research, e-commerce AI and recruitment, yet all pointed to the same change: with AI, businesses are moving beyond traditional measures such as DAU, delivery volume and content output toward measures closer to outcomes. Chuangkit looks at content exports and users’ production efficiency; Jirui Technology has shifted from content delivery to incremental GMV; Xuntu Technology has upgraded from activity metrics to the number of AI task invocations; TTC has moved from résumé counts to first-interview counts; and Nuomibao positions itself over the longer term as “compliance infrastructure.”
A second area of agreement was that AI does not simply replace people: it rewrites service models and organizational structures. Chen Ming discussed how AI combined with people can reorganize legal and tax services. Wu Bin and Yu Chunyan shared how AI improves team productivity per person. Ning Liaoyuan described the opposite, but equally reasonable, situation: AI improves efficiency while opening up demand, so teams do not necessarily shrink. A third consensus was even more important: as AI becomes stronger, human cognition, judgment, taste and broad knowledge become scarcer. Several panelists said the most competitive people will not be those best at executing standardized actions, but those best at defining problems, checking facts and exercising professional judgment.
Next, “Marketing and Sales: AI-Driven Customer Experience Innovation and Conversion Growth” was moderated by Duan Hongyu, partner at Unique Research. Panelists were Wang Zhaoqi, co-founder and COO of Xiaobangbang; Zhang Keyi, founder and CEO of LynxAI; Chen Erhang, senior vice president of Yixuan Technology; and Li Shouguo, CEO of Beta Data.
This discussion clarified a common misunderstanding: AI will not simply replace salespeople; it is more likely to reshape the sales process. The panelists agreed that AI is particularly good at execution and intelligence-oriented work, including prospect discovery, initial screening, alerts, training and workflow coordination. Yet people remain indispensable in high-ticket sales, complex decisions and relationships built on deep trust. Wang Zhaoqi and Zhang Keyi placed greater value on covering the full process: the former already had implementations ranging from customer acquisition to sales practice, while the latter embedded AI capabilities in brand innovation and market insights. Chen Erhang and Li Shouguo took a more pragmatic position: different companies and settings require different entry points. ToB organizations are complex and their data highly fragmented; often, solving one specific problem thoroughly is more valuable than drawing a grand end-to-end loop.
The discussion ended with a simple but important conclusion: the use case comes before the technology. For many businesses, whether AI is worth using depends not on how advanced it is, but on whether it creates clear ROI within existing workflows. Only technology that can actually be implemented deserves to be called a growth tool.
The roundtable “Intelligent Office Work: When AI Takes Over Data Flows, Humans Focus on Creating Value” was moderated by Tang Minglei, co-founder of Panfeng Intelligence and a senior investor. Panelists were Pang Dawei, founder and CEO of Yuankong AI; Gavin Gu Chenggang, founder of ATOA.AI; Ma Liang, CEO of Guiji Geek; and Zhou Ze’an, founder and CEO of Biyou Technology.
This discussion reopened office work as a field that might appear mature. The panelists agreed that 2025 was a year of bringing capabilities into practice and iterating quickly, while 2026 would place greater emphasis on localization, outcomes and applications that cross boundaries. Tang Minglei described the change as a shift from “C (capability)” to “R (results).” A clear view also emerged: upgrades to large models will not swallow every application. Real startup defensibility comes from long-term accumulation within a specific user group, workflow and setting. Speed can create a first-mover advantage, but what ultimately endures is technology, user data and private use cases.
Their advice to founders was direct: build a demo first, establish a small closed loop, collect real data, and only then discuss fundraising and expansion. Investors increasingly value certainty around products, users and revenue. The bar for AI entrepreneurship has not fallen; it now demands that founders validate value faster.
The Next Question for Enterprise Services Is Not Whether to Use AI, but How to Make It Collaborative, Accountable and Implementable
The day’s final main-forum roundtable, “Memory, Security and Collaboration in Enterprise Service Settings,” was moderated by Abner Zhao Liang, partner at Unique Capital. Panelists were Zhao Ming, partner at Future Tense Intelligence; Yang Hongkai, co-founder and COO of Dudao Technology; Liao Can, marketing partner at YUHE Technology; and Jin Lijian, founder and CEO of Yingdao.
This discussion laid out almost all the most practical issues facing enterprise-service AI today. The panelists first acknowledged that enterprise demand for AI has moved beyond “proof of concept”; companies now want to know whether it actually creates value. Zhao Ming focused on quantifiable value with security as a prerequisite. Yang Hongkai emphasized practical results in sales. Liao Can placed “delivery of results” at the heart of AI enterprise services. Jin Lijian, speaking for Yingdao, took a more emphatic position: enterprises should resolutely go All in AI, because this is not optional.
On security, the panelists offered concrete answers from different perspectives rather than speaking in generalities. Zhao Ming proposed treating AI as digital employees when managing permissions and designing workflows. Yang Hongkai emphasized private control of data and traceable processes in heavily regulated industries. Liao Can argued that security in the AI era is fundamentally about precisely defining digital employees’ responsibilities and boundaries. Jin Lijian took a more pragmatic approach: bring AI into the enterprise and start creating value first, then continually address security and governance problems through actual use. On “collaboration,” the differences became more interesting: Jin Lijian even suggested that future organizations might have “AI leading decision-making, with people in supporting roles.” The other panelists tended to understand human–AI collaboration through the division of work, handoffs and risk control.
The closing discussion left the summit with an important conclusion: competition in enterprise-service AI will not be decided by model parameters, but by organizational fit, process design, security governance and delivery of results. Whether AI can enter frontline workflows and genuinely take responsibility is becoming the dividing line in the next phase of enterprise-service competition.
Beyond the Main Forum, Unique DemoDay Brought the Realities of Entrepreneurship Directly to Investors
While the main forum focused on trends, underlying logic and paths forward, the concurrently held Unique DemoDay brought “real market feedback” to the foreground.
For every AI Founder building a startup—especially those preparing to expand overseas—the real challenge has never been simply “having an idea.” It is discovering, upon entering global markets, whether local players have already built barriers; whether compliance, data and localization issues will stop you; and whether technology can actually become growth. Unique DemoDay’s approach was straightforward: bring projects and products before 200+ AI investors and let the market deliver its most candid feedback.
The projects presented that day spanned voice agents, health and eldercare robots, execution-oriented Agents, video creation, agricultural agents, AIGC creation platforms, companion hardware, PC action agents, brand-marketing agents, AI office platforms, Agents for the WeChat ecosystem, social networking, robotics, overseas marketing, video-production engines for the OpenClaw ecosystem, global compliance platforms, industrialized AI-series production, e-commerce business-and-finance Agents, and design tools for the low-altitude economy.
The companies and speakers who took the stage included:
Wei Jiaxing, CEO of Yunfu Intelligence; Zhang Kai, founder and CEO of Xinyi Technology; Gao Jiahui, founder of Luoji Technology; Xu Anbang, founder of Loova; Xu Lianyun, founder of Chunyun Smart Agriculture; Li Kun, founding CEO of Linggan Huabu; Liu Jiaying, founder/CEO of Xingya; Ma Liang, CEO of Guiji Geek; Zhang Keyi, founder and CEO of LynxAI; Li Ju, CEO of Attribuly; Gu Chenggang, founder of ATOA; Li Shiping, CEO of BestClaw; Li Kechen, CEO of Halo; Ma Yao, founder of Nano Robotics; Tony Sun, co-founder of FOSHO; Dennis Zeng Min, CEO of WUI.AI; Zhang Ning of Gloco (Flatfee); Zhang Pinpin, founder of CrunaStudio; Liu Changyu, founder of Qifu Bazhuayu; and Du Kejun, founder of Skyhive AI.
The projects themselves offered a revealing cross-section of AI entrepreneurship today. Some were building intelligent sales platforms already generating sales; others were making health and eldercare robots that can hug and move. Some sought to reinvent the entry point for video creation; others were finding Agent applications in agriculture, cross-border marketing, e-commerce finance, social matching and the low-altitude economy. Some entered the OpenClaw ecosystem; others focused on AI office work and automated PC execution; still others used standardized protocols to reconstruct global compliance services. Together, they illustrated a vivid reality: opportunities in AI entrepreneurship are no longer a single track, but a rapidly expanding industrial map.
A Final Word
From entrepreneurship by super-individuals to the realignment of industry and capital; from medicine, health and eldercare, knowledge management, marketing and sales, and intelligent office work to enterprise services; and from investors updating their judgments to the concentration of startup presentations at DemoDay, one signal from the summit became increasingly clear:
In 2026, AI has fully entered
the stage of “who can turn AI into results.”
Hangzhou is becoming an important testing ground for this new stage. Founders, investors and industry participants are gathering rapidly, while policy, capital, space, talent and ecosystem resources continue to increase. For anyone seeking their next direction, the summit’s greatest value may not have been simply to offer answers, but to make something clearer:
The real opportunities in AI
belong to those who can embed technology in industry
and turn capabilities into outcomes.
And that may be, in 2026,
the industrial question truly worth returning to.
This article was produced by Unique Research.