How Taiwan's AI startup Zeabur secured 64 million in funding, starting from the founder's graduation project.

At 24 years old, entrepreneur Lin Yuanlin from Taoyuan, Taiwan, graduated from the Computer Science Department of Zhejiang University in 2023 and previously interned as a front-end engineer at Ant Group. During his university years, he took on many outsourced development projects but realized that the real time-consuming part wasn't coding but deployment and delivery. This led him to develop a graduation project, which was originally seen as something that would be thrown into a drawer after completion, into an AI startup that spans across Taiwan and the U.S. and has garnered the attention of Silicon Valley venture capitalists. Recently, he also shared his experiences on the YouTube channel Kelly Tsai.

The cloud deployment platform Zeabur (formerly known as Zeebird), founded by him, recently completed a $2 million seed round financing led by the well-known American venture capital firm 500 Global. Currently, the number of registered users has surpassed 100,000, with about 5,000 paying users. The product achieved the top spot on Product Hunt, the world's largest product launch platform, garnering attention in the developer community.

The graduation project unexpectedly became a startup theme, and I was accepted into YC before graduation.

During his university years, he took on outsourced development projects but found that although development was fast, every time he went live, set up the cloud environment, and delivered to the client, he had to research everything again. Therefore, he set his graduation project as an automated deployment platform, trying to enable developers to complete the launch process in a more intuitive way. At that time, he told his thesis advisor, “I want to create a 'backend version of Vercel',” and after creating the MVP, he joined the MiraclePlus startup accelerator.

Originally, it was just a project completed for the sake of fulfilling a requirement, but after practical testing by his roommate, it was deemed actually usable, becoming the starting point for his entry into the entrepreneurial circle. Lin Yuanlin subsequently applied for an accelerator and unexpectedly got accepted into YC China, officially starting his business before graduating.

The biggest misconception of technical founders: Code is not the hardest part.

As a typical technical founder, Lin Yuanlin admitted that he has fallen into many common traps of entrepreneurship. He said, “Many founders with a technical background believe that writing good code equals entrepreneurial success, but in reality, writing code is the easiest part of entrepreneurship.” He pointed out that the real challenges are pricing, sales, business models, and customer segmentation.

Zebra once attempted to compete with platforms like Vercel and Railway, even trying to attract users with free quotas, only to find that most of them were users who would never pay, which instead drained resource allocation. “If your competitive advantage is only cheap or free, it means that the product's value proposition itself is wrong.” This became Zebra's most expensive lesson in its early days.

Two Pivots: From Vibe Coders to Builders

Zebra's positioning has also gone through multiple twists and turns. Initially targeting the engineer demographic, it later shifted towards Vibe Coders who do not know how to program, attempting to ride the wave of AI-generated applications. However, the team only realized after arriving in Silicon Valley that “Vibe Coding” in the American context refers more to informal, non-product-level experimental products, which led founders who genuinely need deployment and infrastructure management to mistakenly believe that Zebra is not suitable for them.

Ultimately, Zebra has redefined its target audience as Builders: regardless of whether they have an engineering background, anyone who is seriously building products and using AI to optimize development and deployment processes is its service target.

AI DevOps Engineer: Can complete assessments and deployments by speaking human language.

In terms of product positioning, Zebra no longer sees itself as merely a PaaS (Platform as a service), but instead focuses on the concept of AI DevOps Engineer. Currently, Zebra's interface is similar to Cursor, but the left side no longer displays code; instead, it shows the user's cloud infrastructure: servers, databases, front-end and back-end service architecture; on the right side, users can converse with the AI Agent through natural language.

For example, users only need to input “Help me buy a cheap server,” and the system will evaluate options such as AWS, GCP, and Linode, completing the configuration and deployment. In addition, the team recently launched Zebra AI Hub, which addresses common model switching and observability issues in AI Agent development, allowing developers to switch between different large models freely without refactoring code and to track token usage and cost sources in real time.

Silicon Valley Startups: Stories Matter More Than Numbers

In order to understand the entrepreneurial culture in the United States, the Zeabur team recently stationed in Silicon Valley for a month. Lin Yuanlin described that this period had a tremendous impact on his entrepreneurial thinking. “Here, entrepreneurship is driven by 'stories.'” He pointed out that compared to Asian investors who pay more attention to existing revenue and data, American venture capitalists often want to hear the vision first and then look at execution capability. He said: “When I pitch in Taiwan, I put the vision on the last page; in the United States, investors instead say that this page should be on the first page.”

He also observed that founders in Silicon Valley are generally extremely optimistic and willing to invest resources very early on to let you try it out. This cultural difference is both a hurdle and an opportunity for Asian entrepreneurs.

Advice for young entrepreneurs: Don't wait until you're ready to take action.

Reflecting on his entrepreneurial journey, Lin Yuanlin offers a seemingly simple yet counterintuitive piece of advice: “Go and do it quickly. Don't let the fear of making mistakes stop you from starting, because not starting is the biggest mistake of all.” In the fast-paced world of AI and software entrepreneurship, he believes that perfectionism can actually be a risk; rapid iteration and quick validation are the keys to survival.

This article discusses how the Taiwan AI startup Zeabur secured 64 million in funding, with the starting point being the founder's graduation project, first appearing in Chain News ABMedia.