Written by: Techub News Compilation

In a16z's podcast, Mintlify's co-founders Han and Hani shared their entrepreneurial journey and how the company evolved from a simple documentation tool into an essential infrastructure in today's AI agent ecosystem. This conversation not only reflected on the hardships and turning points of entrepreneurship but also focused on how the rapid development of AI technology is reshaping product positioning and market demand, providing valuable insights for tech entrepreneurs.

Original Intention and Turning Points: Starting from Solving Their Own Pain Points

Han and Hani founded Mintlify based on their personal experiences as developers. Han recalled self-learning programming since the age of 11, suffering from "bad documentation" during the process. This painful experience motivated them to create a product that truly helps developers. However, the entrepreneurial journey was not smooth sailing. Before finding the direction of "documentation," they went through a year and a half of "wandering in the desert" and at least eight significant pivots.

Hani described that period as a process of "constantly trying many bad ideas." Their initial ideas spanned a wide range, from using AI to automatically describe code functionalities to linking static content with code, but most did not work. This frequent "pivoting" experience was not aimless; its core always revolved around the theme of "helping developers learn and build faster." Multiple failures allowed them to accumulate valuable experience: when an idea truly works, its signal is exceptionally clear.

What ultimately drove them to focus on documentation was the repeated feedback from users that existing documentation tools in the market were not ideal. They realized that instead of believing the misconception of a "saturated market," it was better to return to their original intentions and create a product that they, as developers, genuinely desired. Thus, they quickly built Mintlify's initial prototype over a weekend.

First Customer and "Not Doing Scalable Things"

The initial validation of the product came from a simple yet direct action. At that time, Han's roommate was in the YC incubator, and their company Hyperbeam was transitioning from consumer products to API products. Han showed them Mintlify's prototype: "Hey, check this out, your documentation is much better now." Their response was: "I want to use it now. How do I set it up?"

Han immediately set to help them configure the DNS. This instant, manual service won them their first customer. Starting with the first customer, word of mouth gradually spread, bringing in the second, fifth, tenth, and now thousands of customers. In the early stages, to win users, Han and Hani even promised to manually help customers check and optimize their documentation content, including correcting grammar and improving structure. These "not doing scalable things" created initial customer affection and laid the foundation for later automation tool development.

They once sought advice from Paul Graham on how to handle the scalability challenges posed by this manual service model. The advice received was: "What you're doing now may be what you'll be doing forever." This prompted them to later form a team and develop automated tools, but the founders believe that those "extra efforts" are always crucial in creating customer delight.

Product Evolution in the AI Wave: From Human Documentation to Agent Infrastructure

The explosive development of AI technology has fundamentally changed Mintlify's product positioning and market expectations. Han noted that from the end of 2023 to early 2024, a "tsunami" changed developers' expectations and development approaches. The product, originally just a platform to help humans write and maintain better documentation, has now transformed into an infrastructure serving "humans and AI."

The core of this transformation is that documentation content is not only for human reading but also serves as "fuel" to train and support AI agents and coding agents. As AI agents become commonplace in workflows (for example, 80% of support tickets may be handled by AI agents), the importance of accurate and up-to-date documentation as a "source of information" has risen sharply. Errors in documentation can lead to erroneous outputs from thousands of agents that rely on it, with severe consequences.

Therefore, Mintlify's focus has shifted from the "appearance and experience" of documentation to the "quality and immediacy of the content itself." Han quoted Andrew Kaparthy's viewpoint: "English will become the hottest programming language," emphasizing the core position of content. While the design of the product interface is important, today, AI agents access documentation through code or raw HTML/Markdown, and they do not care at all about "aesthetics." The future focus will be on ensuring content excellence and real-time updates.

Self-Revising Documentation: A Long-term Problem About to be Solved

Han analyzed that there are several reasons why documentation has been difficult to maintain for a long time: Organizational Dynamics: The ones who understand the functionality and have the context best (usually engineers) are not the ones responsible for maintaining the documentation; Human Factors: People generally dislike updating documentation, viewing it as a tedious task.

Now, the combination of three factors makes "self-updating, self-repairing documentation" likely to be achieved: Urgent Demand: The impact of outdated documentation on AI agent workflows is real-time and serious; Mature Model Capabilities: For instance, models like Opus 4.5 are now reliable and consistent enough to handle this task; Trust and Environment: Companies are now more willing to provide contextual information to LLM agents, which was unimaginable two years ago.

This means that the long-standing issue of "knowledge content becoming outdated," which has existed for the past 25-30 years, may finally be resolved. Mintlify is building tools to realize this vision.

Customers and Market: Insights Gained from Serving the Most Demanding Customers

Although Mintlify is a young company, its customer list includes top tech companies like Anthropic, Microsoft, and Coinbase. Han candidly said that collaborating with these "demanding and fast-moving" companies is a privilege but also brings immense challenges and learning opportunities.

The response speed of these companies is astonishing. Whenever a Slack message is sent, a reply is almost guaranteed within 10 seconds. This work pace motivates the Mintlify team to hold themselves to the same high standards. Han recalled an instance when Anthropic's documentation was suddenly auto-translated from English into twelve other languages; they were surprised and asked if a translation team had been employed, to which the response was: "We built a real-time translation pipeline." This capability, which was hard to imagine two years ago, has now become the norm and inspired Mintlify's own product innovation.

Collaboration with cutting-edge companies has not only brought product inspiration but has also changed their understanding of market demand. Today, Mintlify's usage scenarios have greatly expanded, extending from public documentation to internal knowledge bases and help centers, serving a broader user base, including engineering teams and HR teams. The product is evolving from merely "documentation" to a broader concept of "knowledge management."

Future Outlook: Empowering Builders and Expanding Worldviews

Looking ahead, what excites Han the most is still Mintlify's original intention: empowering builders. He mentioned that last month, 20 million people visited sites powered by Mintlify. Behind this may be many individuals, like himself at age 11, who are learning programming for the first time and utilizing documentation information to build things for their community or school. This broad impact serves as the driving force behind his work.

Hani is excited about two things: first, building more AI agents and constantly being amazed by their emerging capabilities; second, as the company grows and their worldview expands, solving problems for more communities, such as now starting to serve customer support sectors.

Both founders believe that AI is merely a powerful new tool that should be openly embraced and used to help others and create value. In the face of rapid technological changes, they maintain a pragmatic approach: on one hand, focusing on macro trends, and on the other hand, addressing users' actual issues, striking a balance between "grand visions" and "mundane yet necessary product decisions."

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。