Before establishing Extreme Vision, a prominent artificial intelligence (AI) company in China, Macao entrepreneur Chan Chan Kit had already learnt one of the toughest lessons in business: failure is the best teacher.
He and his partners failed twice prior to their success with Extreme Vision. Their first venture – a job-search platform for blue-collar workers in the Chinese mainland – emerged while Mr Chan, who turns 34 this year, was still a university student. The second was closer to home, an attempt to digitise restaurant ordering and delivery in Macao. Neither venture succeeded. Yet those early setbacks left the team with a sharper understanding of what truly matters when building a technology company: demand matters more than novelty, and hype is no substitute for good timing.
Those lessons shaped the strategy Mr Chan and his partners would later apply to Extreme Vision, a company specialising in AI systems. Rather than trying to outrun larger and better-funded competitors in an already crowded field, they chose to focus on a market they believed would take years to mature. “We’re just ordinary young people without abundant resources, so we have to choose industries that may only take off in five to 10 years, giving us time to trade for space to grow and expand the business,” he remarks.
They set their sights on AI more than a decade ago, long before it became today’s global buzzword. Shenzhen Extreme Vision Technology Co., Ltd. was established in 2015. Over the following years, the business expanded beyond its original base, relocating its headquarters to Shandong province in 2021 and rebranding as Shandong Extreme Vision Technology Co., Ltd. Earlier this year, the company reached another milestone by listing on the main board of the Hong Kong Stock Exchange, having already served more than 3,000 clients across the Chinese mainland, Macao and Hong Kong.
Models for success
Currently, AI computer vision solutions and large model systems form the company’s two core business lines. The first uses AI to interpret and analyse images and videos, extracting information that can be applied to real-world situations. “Our eyes can identify things, for example, who a person is, which is facial recognition. They can also tell whether there’s a fire at a site, which is what we call fire detection,” he says in an interview with Macao Magazine. “Put simply, what AI computer vision solutions do is automate the work that the human eye does.”
The technology already has a wide range of practical applications, from automated quality inspection and object detection to facial recognition and hazard monitoring. Traffic enforcement provides one of the clearest examples of computer vision deployed at scale. Such systems can identify speeding, detect seatbelt non-compliance, flag mobile phone use while driving, and spot failures to yield to pedestrians, wrong-way driving and lane guidance violations.
Extreme Vision’s second growth engine is newer, but equally ambitious. Its large model solutions are built around large language models and multimodal models that help organisations automate complex workflows, generate content and extract insights from vast datasets, improving productivity, customer service and decision-making.
“Our large model solutions serve as the underlying AI operating system for enterprises,” Extreme Vision’s chairman and general manager says. “The trend for the future is that a company may no longer have 10,000 employees. It could be 1,000 people plus 9,000 agents, including AI or machines.”
The challenge, he argues, lies in bringing those people and AI agents together effectively. “The key is how 1,000 people and 9,000 agents collaborate, how workflows are stitched together, and how system security is built in. That’s what our large model solutions are designed for: a future in which companies will have large numbers of humans working alongside machines.”
For Mr Chan, the technology itself is no longer the biggest hurdle. As AI becomes increasingly accessible, success depends less on developing the underlying models and more on ensuring they deliver measurable value to customers. That philosophy has become a defining feature of Extreme Vision’s approach as it seeks to expand both within China and overseas.
Value creation and hybrid talent
Extreme Vision’s listing prospectus illustrates the scale the company has achieved. As of September 30, 2025, it had completed more than 6,000 projects since its establishment and recorded a product repurchase rate of more than 80 percent. It also operates an AI computer vision algorithm marketplace, enabling enterprises to select from an extensive library of algorithms. By September 2025, the marketplace featured 1,517 algorithms, including 148 developed in-house and 1,369 co-developed with third-party developers. Together, they cover more than 100 industries, ranging from smart transportation and urban management to retail, utilities and manufacturing.
Despite the breadth of applications, Mr Chan believes AI’s greatest potential still lies ahead. “From a practical standpoint, I don’t think there is any industry that doesn’t and won’t need AI. The difference is only how deeply it’s used and how much potential it unlocks,” he notes.
He points to examples in the Chinese mainland, where local governments have introduced AI-powered tools on public-facing websites to answer citizens’ enquiries, often identifying the relevant regulations and guidance more quickly – and more consistently – than a human operator could. “At its core, AI is a supplement to human work and, to some extent, a substitute,” he adds. “As long as an industry relies on people, it will need AI.”
Of the more than 3,000 clients the company has served, Mr Chan says over 95 percent are based in the Chinese mainland. They include government departments, state-owned enterprises and private companies. Most of the remaining clients are in Macao and Hong Kong, often involving projects for public-sector organisations. Working across different jurisdictions, he explains, requires the company to continually refine its algorithms to meet varying regulations, operational requirements and enforcement practices.
As the AI industry has evolved, so too have the challenges of building an AI business. In the company’s early years, technological capability was the biggest constraint. Mr Chan now believes the real challenge lies elsewhere: creating value for clients. “The core issue is that AI is meant to empower industry, but AI teams don’t understand the industry and industry teams don’t understand AI, leaving a gap to be filled,” he illustrates.
Bridging that divide, he argues, requires a new type of professional with expertise in both technology and business. “We need more consultant-type talent and industry-integration talent to fill that gap. Then companies will be enabled to truly use AI as a tool to lift productivity, calculate internal returns and other purposes,” he says.
To address that need, Extreme Vision has increasingly recruited industry specialists and consultants before training them in AI technologies and placing them in customer-facing roles, where they can help clients identify practical applications rather than simply introducing new technology.
From Macao to Qingdao
Extreme Vision’s three executive directors – Mr Chan and his partners Luo Yan and Chen Shuo – have been friends since their university days at Sun Yat-sen University in neighbouring Guangdong province. Each brought a different area of expertise to the business. Mr Chan studied business, completing a bachelor’s degree in supply chain management before earning a master’s degree in enterprise management from Peking University. The other two partners, meanwhile, built careers spanning computer science, mathematical statistics and biotechnology.
Their entrepreneurial journey reflects a broader wave of innovation that emerged across China during the 2010s, encouraged by national policies promoting mass entrepreneurship and innovation. Even while at university, the three friends believed technology could provide an opportunity to build something with lasting value. After two unsuccessful ventures, they turned their attention to AI, initially exploring opportunities in Macao before formally establishing Extreme Vision in Shenzhen more than 10 years ago.
For Mr Chan, beginning in Macao was both a personal and strategic decision. Born and raised in the city, he was naturally drawn to launching the business there. Macao also offered valuable research resources through its universities, while its client base – despite the city’s relatively small size – included government departments, integrated resort operators, banks and higher education institutions that could serve as early adopters of emerging AI technologies.
As the company grew, however, the founders realised that building an AI business requires substantial upfront investment, together with access to specialised talent. Financing and recruitment quickly emerged as significant challenges that were difficult to overcome in Macao at the time.
That reality gradually shifted the company’s centre of gravity to Shenzhen, where a more mature venture capital ecosystem and a deeper talent pool provided the resources needed to support rapid expansion. In Mr Chan’s view, Shenzhen’s greatest strength lies not simply in the size of its market, but in the complete financing ladder available to technology companies, from early-stage investment through to later funding rounds, as well as its concentration of highly specialised technical professionals.
The company’s next move took it further north, relocating its headquarters to Qingdao in Shandong province in 2021 and adopting its current corporate identity. Mr Chan sees Shandong as an ideal base for an industrial AI company. As one of China’s largest provincial economies, it has a broad and diversified industrial base that includes chemicals, coal, advanced manufacturing and numerous other sectors.
Since 60 to 70 percent of Extreme Vision’s clients come from industrial areas, Shandong is an ideal environment for developing AI applications at scale before replicating successful solutions across the rest of the country.
Overseas expansion
The company’s growth has been striking, as per its own disclosures. According to figures included in its listing prospectus, revenue more than doubled from 101.6 million yuan in 2022 to 257.3 million yuan in 2024. Revenue also increased by 71.7 percent, from 79.4 million yuan in the first nine months of 2024 to 136.3 million yuan during the corresponding period in 2025.
Its Hong Kong initial public offering (IPO) in March raised 499 million Hong Kong dollars. The proceeds have been earmarked to strengthen research and development, recruit top-tier multidisciplinary talent, acquire critical technical services, expand its marketing network and accelerate overseas expansion.
Mr Chan says the company’s immediate priority is to strengthen both of its core businesses – AI computer vision and large model solutions – while helping more enterprises integrate AI into their operations.
“We hope that, after further optimising our operations, we can help more Fortune Global 500 and China’s top 500 companies build the next generation of AI underlying operating systems,” the company’s chairman states.
International expansion is now the next major milestone. Mr Chan says the company has already held discussions with potential partners and customers in markets including Thailand, Indonesia, South Korea and Japan.
With a track record built in the Chinese mainland, and with China and the United States widely regarded as the world’s two leading AI powers, he believes the company is well positioned to compete internationally.
“We’ve come out of the Chinese market, one of the world’s most competitive environments,” he notes. “In terms of product competitiveness, Chinese tech products today are extremely strong. Compared with the U.S., the technology can be similar, but the cost can be as low as one-tenth.”
Importance of Macao identity
Alongside the company’s rapid growth, Mr Chan has also received recognition for his own entrepreneurial achievements. He has been named on the Forbes “30 Under 30 Asia” list and on Hurun China’s “Under 30s to Watch” list.
Despite leading a company whose operations now stretch beyond Macao, Mr Chan says the city continues to shape both his identity and the way he does business. He argues that local governments on the mainland pay close attention to companies founded by entrepreneurs from Macao and Hong Kong, often maintaining regular contact to understand the challenges they face and identify ways to offer policy support. He also points to a range of initiatives designed to encourage Macao residents and businesses to establish themselves beyond the city.
For that reason, he encourages young entrepreneurs in Macao to think beyond the city’s boundaries. The mainland, he explains, offers a dense network of incubators and start-up platforms that enable founders to build entrepreneurial networks far more quickly, transforming introductions into partnerships, as well as providing recruitment opportunities and new business ventures.
Macao’s own start-up ecosystem has improved significantly in recent years, he acknowledges, but it cannot match the mainland’s market size or depth of demand.
“If you want to develop a technology business, tapping the mainland market is an inevitable step,” he points out. “One defining feature of tech start-ups is the heavy upfront investment. Before listing, we had already raised a cumulative 400 million to 500 million yuan. From starting from scratch to reaching break-even, we invested at least 300 million yuan. That kind of investment is hard to recoup if the business is limited to Macao.”
Trailblazer over trend chaser
Mr Chan’s final piece of advice to aspiring technology entrepreneurs is simple: don’t chase whatever happens to be fashionable. Trendy sectors inevitably attract both capital and competition. By the time a field becomes the latest investment craze, it is often already crowded with well-funded market leaders, making it far more difficult for newcomers to establish themselves.
Instead, he believes entrepreneurs should identify industries with long-term potential, areas that may still be underdeveloped today but are likely to become mainstream over the next decade.
To illustrate the point, he quotes a line widely attributed to Lei Jun, the Chinese billionaire entrepreneur and founder of Xiaomi: “Stand in the path of the wind, and even a pig can fly.” Beneath the humour lies an important lesson: entrepreneurs who recognise where an industry is heading – and position themselves early – can benefit from powerful long-term momentum.
“It’s hard for new start-ups to break through in crowded, popular sectors,” Mr Chan stresses. “But areas likely to grow over the next decade, such as AI in healthcare and AI in genetic engineering, still have relatively low penetration and fewer entrenched players, which can translate to more attractive opportunities.
“We need to seize the moment as these sectors move from niche to mainstream. That’s where the opportunities and growth will be,” he highlights.
