The former youngest vice president of DingTalk has gone all-in on AI Agent entrepreneurship.

Wang Ming, a former senior executive of DingTalk who once sat at one end of the round - table and had a fierce debate with Zhu Xiaohu about "whether large models can really make money", has now given his answer with practical actions: he has personally embarked on an entrepreneurial journey.

In Hangzhou in January, the sun still shines brightly. Sitting in an office by the Beijing - Hangzhou Grand Canal, bathed in the warm winter sunlight, Wang Ming looked out of the window. Just one or two years ago at the DingTalk AI Ecosystem Conference, he firmly asserted: "In AI entrepreneurship, one should firmly choose the To B model! The key to making money lies in validating customer value. Even if the initial revenue is limited, as long as you can get real feedback and continuously optimize the product, you can stay in the game. When the cost drops and the model stabilizes, you can enter the profit - making stage after productizing the solution."

Now, Wang Ming has stepped out of his former role as a "pragmatic senior executive" and started a new journey in this new office - an AI Agent entrepreneurship targeting overseas C - end Prosumers (professional consumers). "I left the big company because I want to be a front - line practitioner in the AI wave, constantly learn and create, and build a real AI partner that can complete tasks," he said.

On January 13th, K2 Lab, founded by Wang Ming, officially announced the completion of a seed - round financing of tens of millions of yuan. This round of financing was exclusively invested by Yunshi Capital. According to Touzhongwang, this team is fully committed to polishing the world's first AI Agent that can directly help creators make money.

The "Entrepreneurship" of the Youngest Vice - President of DingTalk

As the youngest vice - president of DingTalk, Wang Ming's decision to start a business stems from a simple belief: the ultimate value of technology does not lie in showing off skills, but in whether it can truly serve people.

In his view, the overabundance of AI - generated content is triggering a serious trust crisis, and AI technology is profoundly reshaping the production relations. "In this transformation, the personified and character - set 'Super Individuals' will become the new key nodes in business. Their demand will experience explosive growth, but their pain points in content creation, commercial monetization, etc. have not been fully addressed, which contains huge market opportunities."

Therefore, Wang Ming believes that for entrepreneurs, the real AI opportunities not only lie in the large models themselves but also in seizing the new opportunities brought about by the economic transformation in the new era: how to make AI "affordable, accessible, and useful" for ordinary people. Instead of turning people into operators of AI, we should make AI an extension and substitute of people, transforming people's lifestyles, personal charm, aesthetic taste, and professional knowledge into online IPs, which can become the anchor points of trust.

Meanwhile, his years of experience in big companies have given him sufficient systematic insights and the confidence to implement ideas. As a reserve cadre in the Alibaba system, he once led the innovation of DingTalk's AI products, including the construction of AI applications and large - model ecosystems, pioneered several intelligent integrated products, achieved nearly tens of billions of yuan in ecological commercialization results, and built China's largest enterprise - service ecosystem.

Especially after the fierce debate with Zhu Xiaohu, Fu Sheng and others at the DingTalk AI Ecosystem Conference, Wang Ming became more aware that it is difficult to reach a consensus, but actions can be taken. At that time, Zhu Xiaohu believed that large - model APIs would eventually be free, just like "water, electricity, and gas", especially in China. Entrepreneurs should focus on "delivering services with AI" rather than selling models or software, and achieve a commercial closed - loop through the "AI + human" hybrid model.

Wang Ming, on the other hand, believes that large - model manufacturers should "figure out what they are for" and must be bound to application scenarios. Servitization and solution - orientation are the directions. In his view, in the AI era, the platform - ecosystem collaboration of applications is the king, while Zhu Xiaohu prefers the "human + AI" hybrid delivery model.

Even earlier, Wang Ming successfully led the incubation of Quick Dog Taxi and Swan Home at 58 Group and founded 58 Enterprise Services and other star products, accumulating complete experience in building platform - type products from scratch and achieving commercial implementation.

These experiences enabled him to acutely capture the structural opportunities in the AI era: First, overseas AI products such as Jasper, Synthesia, and HeyGen have achieved annual revenues of hundreds of millions of dollars, validating the willingness to pay for "AI as a service". In China, although the enthusiasm for entrepreneurship is high, most teams are still at the Demo stage and lack the ability to transform technology into sustainable revenue. Second, although big companies have advantages in computing power and data, they are constrained by organizational inertia and KPI - path dependence and find it difficult to quickly test and iterate in segmented and rapidly changing AI scenarios, which is exactly a golden opportunity for startups.

Based on this, in October 2025, Wang Ming and several core partners decided to step out of their comfort zones and officially founded K2 Lab. The company positions itself as the "Shopify of the AI era", focusing on the Agent infrastructure for "Super Individuals" and is committed to building a commercial video Agent that can complete end - to - end content e - commerce understanding, industry insights, topic planning, script generation, multilingual dubbing, video editing, platform distribution, effect review, fan operation, and even supply - chain management.

Because Wang Ming believes that the ultimate goal of influencer commercial marketing is a multi - modal AI Shopify.

Under this strategy, the team has quickly assembled members with diverse backgrounds, including former DingTalk technical backbones, senior overseas growth experts, interaction innovation experts, and star content e - commerce operators, to jointly participate in this wave of AI Agent application entrepreneurship.

Reaching the "End - Game" of Commercial Marketing through Four Differentiations

After choosing the track, the feasibility of the market and product becomes the key to implementation.

Wang Ming said that K2 Lab's strategy is to first lock in the core track: starting with the commercial monetization of overseas creators, focusing on developed countries in Europe and America, and taking the United States as the first market for implementation.

Why choose Europe and America? This is based on three key judgments: First, the creator groups in the European and American markets have a mature willingness to pay, and content e - commerce monetization is still in a blue - ocean market, with concentrated and urgent pain points. Second, platforms like TikTok have accumulated 2 billion monthly active users, but the supply of commercial content is seriously insufficient, and there is a huge room for improvement in the monetization rate. Third, Chinese teams have natural differential advantages in e - commerce operation and AI application implementation in overseas markets.

Different from most AI tools in the market, Wang Ming told Touzhongwang that K2 Lab's core product, Mora, breaks out of the mindset of "functional tools" and builds its core competitiveness with a "result - oriented" approach. Its differential advantages are reflected in four dimensions - model difference, interaction difference, technology difference, and payment difference.

Among them, the model difference means abandoning the single - tool logic and providing a full - link solution from product selection, script creation, video generation to account operation and data analysis, directly delivering the "commercial monetization result" rather than the "tool usage permission". The interaction difference takes the human - like Agent as the core interaction form, minimizing the participation threshold for creators. Creators can complete content creation and monetization without complex operations, and at the same time, they can strengthen user trust by continuously interacting and precipitating their personas.

The payment difference breaks the traditional subscription - based model and adopts a result - based payment mechanism of "basic salary + commission". Creators only need to pay a small basic salary and can enjoy the commercial revenue share brought by AI, achieving "low - threshold entry and high - return rewards". The technology difference is that K2 Lab uses a Multi - Agent architecture, integrating the capabilities of nearly 20 mainstream large models, covering core technologies such as reasoning, large - language processing, multi - modal understanding, and generation. Through the collaborative cooperation of product - selection Agents, creation Agents, operation Agents, and analysis Agents, the whole process can run automatically.

Most importantly, although K2 Lab has only been established for two months, its key product has completed preliminary verification: it can generate a promotional video that meets the tone of TikTok in 30 seconds, helping creators lacking creation or product - selection ability to quickly run through the business model, and the computing - power cost can be effectively covered by commercial monetization.

Many Senior Executives from DingTalk Have "Dived into Entrepreneurship"

In this year of the explosion of AI applications, Wang Ming is not the only one who has started a business. A noteworthy phenomenon is that a group of core senior executives who were deeply involved in DingTalk's strategic transformation from a tool to an ecosystem and then to AI are intensively leaving big companies and diving into the wave of AI - native entrepreneurship.

Besides Wang Ming, other senior executives from DingTalk are also quietly deploying in the AI Agent track. For example, Lin Feng, the former vice - president of DingTalk and the general manager of the intelligent hardware ecosystem, plans to start a business after leaving in 2025, focusing on the field of AI office hardware. As a veteran technical expert in Alibaba, he once led innovative projects such as DingTalk's XR hybrid office and VR/AR payment. Moreover, Ye Jun, the former CEO of DingTalk, also left in July 2025 and entered the AI Agent track, focusing on the development of intelligent agents for To B scenarios.

Together with Wang Ming, these entrepreneurs generally have three commonalities: First, they were deeply involved in the implementation of DingTalk's "All in AI" strategy. Second, they are familiar with the cross - scenarios of B - end and C - end. Third, they have first - hand verification experience in "how AI can truly improve productivity".

Now, their unanimous focus on the AI Agent track further shows the huge potential and structural opportunities in this track.

Regarding the future, Wang Ming said that K2 Lab has formulated a clear three - dimensional plan of "product iteration and growth + financing promotion + technology upgrade".

In terms of product iteration, K2 Lab will complete the polishing and ROI calculation of various channels in the US market from March to April, start large - scale promotion, and quickly increase the penetration rate. In the second half of the year, functions such as interactive fan - growing, fan private - domain management, emotional - value services, and supply - chain - ability docking will be gradually launched to enrich the product matrix, and interaction innovation combined with hardware will be explored to further lower the user threshold.

In terms of technology upgrade, in the first half of the year, the product experience will be optimized based on the existing large models. In the second half of the year, as the user scale and data accumulate, K2 Lab will start model fine - tuning and post - training to improve the AI basic capabilities in vertical scenarios and solve more core pain points.

In terms of the financing rhythm, after the completion of the first - round financing, Wang Ming revealed that a new round of financing will be completed before the Spring Festival. After June, on the basis of the rapid growth of data and business indicators, a larger - scale financing will be launched. "In short, throughout 2026, we plan to accumulate a capital reserve of tens of millions to 100 million US dollars to support technology R & D and global market expansion."

"We believe that the entrepreneurial competition in the AI era is faster and more intense than ever," Wang Ming emphasized. "But it is precisely such an opportunity that makes us not stay in the hustle and bustle of chasing technology but return to the needs themselves - listen to those real voices and serve them."

Finally, as the conversation was coming to an end, the sun outside the window was still bright. In this small office, a "transformation" about individuals, technology, and the future of global business is quietly accelerating. This time, Wang Ming and his team will fight together with the "Super Individuals".

This article is from the WeChat official account "Dongshisi Tiao Capital" (ID: DsstCapital), written by Chen Mei and edited by Wang Qingwu, and is published by 36Kr with authorization.