At the age of 19, she raised 120 million yuan in financing.

“I just turned 19 recently. I dropped out of college's computer science major to start a business, aiming to create high-quality code data to support large AI models.”

These words come from Serena Ge, a post-2000s girl with a Chinese face. Now, she is the co-founder and CEO of the AI data company Datacurve. The company has only been established for one year, but it has already secured $17.7 million in financing (equivalent to approximately 126 million RMB).

Quietly, an AI entrepreneurship wave among the post-2000s generation is sweeping across the globe.

19-Year-Old Girl Starts a Business in College

AI "Shovel Sellers"

In 2006, Serena Ge was born in China. Later, she moved to Canada with her parents and had the idea of starting a business at a very young age.

In high school, due to her love for rock climbing, she developed a personalized rock climbing training app, which was well-received by rock climbing enthusiasts. After that, she led a team of 23 high school students to jointly develop an efficiency web application for teenagers, and the project was supported by the Bank of Montreal in Canada.

In 2022, she was admitted to the University of Waterloo in Canada to study computer science. However, Serena soon found that the mainstream atmosphere on campus was to "find a decent and stable job after graduation," which was completely inconsistent with her desire to engage in cutting-edge technologies and create the future with her own hands.

The turning point came in 2024. At that time, with the AI agent project UncleGPT, she received an invitation from the startup incubator Y Combinator. Almost without hesitation, she chose to drop out of school. Also in this year, Serena and her Asian classmate Charley Lee jointly founded the AI data company Datacurve.

This entrepreneurship originated from her internship experience at the large AI model unicorn Cohere, where she worked as a machine learning engineer. During that time, she found that it was difficult for peers to obtain expert-level annotated data. Limited by costs and other reasons, AI annotation companies would not recruit high-quality software engineers to do the most basic data annotation work.

"The bottleneck of large models lies in the lack of rich, carefully selected, and high-quality annotated data," Serena said. This is exactly the data problem that Datacurve wants to solve.

Different from Scale AI, which relies on a large outsourcing team, Datacurve's data collection model is quite interesting. It uses a "bounty hunter" system to attract skilled software engineers to complete the most difficult part of data acquisition.

Put simply, on Datacurve's "bounty platform" Shipd, more than 1,400 programmers take on challenges in tasks such as algorithms, testing, and UI/UX processes. For each task solved, users can receive a fee ranging from $5 to $50. This incentive mechanism aims to reward quality and speed. So far, the company has paid out more than $1 million in bounties.

"This is a user-oriented product, not just a simple data annotation job," Serena said. The biggest motivation for engineers to participate is not money. In fact, the remuneration for data annotation is always lower than that for services such as software development. Therefore, the company's core competitiveness lies in providing a good user experience and attracting more high-quality programmers to join.

The company said that as large language models continue to evolve, artificial intelligence no longer only needs simple data annotation, but a large amount of training data and evaluation data. The company uses a gamified platform to improve the accuracy of data generation and marking, and then complete high-quality data delivery.

Currently, the team only has about 10 people and is still recruiting. The company said that its revenue exceeded $1 million two months after its establishment. Now, it has provided high-quality code data for more than half of the basic model laboratories and companies such as Facebook, Apple, Amazon, and Google, helping to train the next generation of more advanced large language models.

Just Secured 100 Million in Financing

It was not until the latest financing that Datacurve caught the attention of the venture capital circle.

Recently, the company completed a $15 million Series A financing, led by the venture capital firm Chemistry VC. Other investors include Y Combinator, Afore Capital, Homebrew, and investors from companies such as DeepMind, OpenAI, Anthropic, Vercel, and Coinbase.

"This is one of the fastest-growing startups we've ever invested in. Just last week, Datacurve signed its largest contract to date," recalled Mark Goldberg, a partner at Chemistry, who remembered Serena's hard work when he first met her.

Even earlier, the company completed a $2.7 million seed round of financing, supported by institutions such as Y Combinator, Afore Capital, and Pioneer Fund. Balaji Srinivasan, the former chief technology officer of Coinbase, participated in the investment.

So far, in just one year, this post-2000s founder team has raised a total of $17.7 million, equivalent to approximately 126 million RMB.

There is a well-known joke in the AI circle: "The more human effort, the more intelligence." Most data annotation companies have large outsourcing teams to refine data and are also jokingly called "Cyber Foxconn." But to some extent, data annotation companies focus on the most essential part of AI: no matter how technology evolves, model training always depends on "clean" data, which is the fundamental reason why data annotation is irreplaceable.

To date, data, algorithms, and computing power are the three cornerstones of AI. If NVIDIA is the "shovel seller" for computing power, then data annotation companies are the "shovel sellers" for data.

In comparison, her competitor is more well-known - Scale AI. In June this year, Meta invested approximately $15 billion, and Scale AI's valuation exceeded $29 billion at once.

Here, we have to mention Edwin Chen, also of Chinese descent. The Series A financing of his company, Surge AI, is seeking $1 billion, and its corresponding valuation has risen to approximately $24 billion (equivalent to about 171.2 billion RMB). Since he holds about 75% of the company's shares, his net worth has reached $18 billion. He made his debut on the Forbes list of the richest Americans and became the youngest billionaire this year.

"This is just the beginning. We will use this funds to accelerate the development of basic models - by providing cutting-edge training data for large language models, we will push the boundaries of AI capabilities," Serena said, adding that she firmly believes that the progress of AI is not only limited by computing power but also by the quality and complexity of data.

The 00s Generation is Dominating the AI World

Unconsciously, it is no longer a legend for post-2000s entrepreneurs to raise hundreds of millions in financing. It happens almost every day.

Just last week, Axiom Math officially completed its first round of financing of $64 million (equivalent to approximately 460 million RMB), led by B Capital, with participation from institutions such as Greycroft, Madrona, and Menlo Ventures. After the investment, its valuation reached $300 million (equivalent to approximately 2 billion RMB).

The founder of Axiom is Carina Hong, a post-2000s entrepreneur. Born and raised in Guangzhou, she attended the prestigious South China Normal University Affiliated High School and won multiple medals in the Olympiad competitions.

Not long ago, two post-2000s entrepreneurs from MIT - 22-year-old Chinese-American Jessica Wu and 23-year-old Neil Deshmukh - founded Sola Solutions and secured financing from well-known venture capital firms in Silicon Valley. According to the official website, the financing includes a $3.5 million seed round led by Conviction and a $17.5 million Series A led by a16z with participation from Conviction, totaling $21 million (equivalent to approximately 150 million RMB).

There is also the AI annotation company Mercor, which is seeking a valuation of approximately $10 billion. The company has completed two rounds of financing, and its valuation after the Series B financing is approximately $2 billion. The founders of the company are three post-2000s entrepreneurs who retired from their studies. In their sophomore year, they founded Mercor in their dormitory and later decided to drop out of Harvard and Georgetown University to focus on their business full-time.

Similar stories are also happening in China.

At the beginning of this year, three post-2000s "geeks" from Tsinghua University - Min Yuheng, Cheng Yi, and Li Yizhe - started a business in robotics, and Lingfang was born. Now, the company has successfully completed angel + and angel ++ rounds of financing in the hundreds of millions, attracting well-known institutions such as Jijie Capital, Tongchuang Weiye, Lihe Technology, Shuimu Fund, Mizuho Lihe, and Ralph Ventures.

We can see that more and more young faces are emerging, such as Chen Yuanpei, the co-founder of Lingchu Intelligence, Yang Fengyu, the founder and CEO of UniX AI, and Jiang Zhenghao, the founder of Chongsun Technology.

Time waits for no one. "AI doesn't wait. If you're one step behind, you'll really miss the opportunity," said these founders, most of whom started writing code at a young age. They believe that artificial intelligence is a once-in-a-lifetime opportunity. Some have given up their degrees from prestigious universities to start businesses, and some have even given up the chance to go to college to pursue their dreams in the AI field.

These young people carry many amazing labels: geniuses, dropouts, hard workers, geeks, one-person companies... It's no longer rare to see a 20-year-old CEO of a unicorn company, and there are many post-2000s entrepreneurs leading teams of dozens or even hundreds of people. As a Silicon Valley investor said, "Starting a business at 19 is not considered early these days." They work out at night, write code during the day, negotiate financing at noon, and launch demos at night - this is the daily life of Generation Z entrepreneurs. They are reshaping the future of AI at an amazing speed.

Liu Yuan, a partner at ZhenFund, once pointed out that "AI has redefined the threshold for entrepreneurship. Technology has put everyone on the same starting line, and the biggest advantage of this generation of young people is that they have no baggage, learn quickly, and act even faster."

The waves of the new generation are surging. This is the era opportunity for this generation.

This article is from the WeChat official account “Touzizhijie” (ID: pedaily2012). Author: Wang Lu. Republished by 36Kr with permission.