I've been a product manager at one of China's biggest tech firms. Here's how Chinese AI products are built differently.

• A product manager who worked at a Chinese tech giant says Chinese AI products differ from the West.
• Chinese and overseas markets are fundamentally different, said Yilin Zhang, who worked at Meituan.
• Zhang, who moved to an AI startup, said more people are choosing entrepreneurship over traditional paths.

This as-told-to essay is based on a conversation with Yilin Zhang, an AI product manager at AI startup Kuse who worked at Meituan for more than three years. It has been edited for length and clarity. Business Insider has verified his employment and academic history.

I graduated from Tsinghua University with a master's degree in computer science in 2021 and then joined Meituan — one of China's biggest tech firms — as a product manager.

At Meituan, China's platform for local services, especially known for food delivery, I worked on two AI projects. One was a consumer-facing AI assistant that helps users complete various tasks, including ordering food. The other was a merchant-facing AI agent designed to help businesses manage their daily operations, including handling reservations, managing orders, and supporting routine operational tasks.

The main difference between how products are built in China and in the US comes down to the market.

Why Chinese tech companies are so cost-efficient

Across most large Chinese tech companies, AI product development accelerated more aggressively around 2025.

The AI initiatives I worked on at Meituan started around April or May of that year. It coincided with the surge of interest around DeepSeek, when attention around AI agents took off.

Large companies began racing to build AI projects, and almost every business unit launched its AI initiative.

For a long time, especially before 2021 or 2022, Chinese tech companies were primarily focused on domestic competition rather than overseas expansion. Because competition in China is intense, tech companies were forced to become extremely efficient. Their execution methods have been sharpened to an almost frightening degree.

Constraints have also pushed Chinese AI companies to pursue different paths, with a strong focus on open-source models and cost efficiency. These limitations forced exploration in new directions, and those paths have proven valuable in their own way.

DeepSeek is a good example. Because of international restrictions, it couldn't access large numbers of GPUs and was forced to innovate around efficiency instead.

Why Chinese AI products differ from the West

Chinese and overseas markets are fundamentally different, leading to distinct user bases, expectations, and product designs.

Chinese users have a much lower willingness to pay for software; hence, many mass-market AI products, such as Doubao, tend to be free. The core objective is often to scale active usage.

Many capabilities are packaged into a single prompt you can ask, essentially a chatbox interface with a low barrier to entry.

International AI products target users doing high-value tasks. They are more often designed for desktops than for mobile devices, with interfaces better suited to work contexts. These products explore how AI and humans can collaborate and intersect across different work scenarios, helping users complete tasks more effectively and efficiently.

In China, that user group is relatively small. That makes it harder for its mainstream AI products to move beyond chat-based forms into more advanced products.

China's internet success over the past decade has also largely come from consumer-facing apps. That environment forces product managers to obsess over user feedback and relentlessly polish even the smallest features.

Teams may spend enormous effort refining a tiny feature just to win over a small group of users. In markets with less competition, that level of detail isn't always necessary.

The AI startup scene is growing in China

After three to four years at Meituan, I felt I had learned most of what I could from that environment. I left to join the AI startup Kuse in October.

AI is evolving extremely fast. In large companies, iteration speed can be slower. Many of my friends across different Big Tech companies share this same frustration. Smaller, more agile companies can adapt faster.

In the past, top graduates had basically two paths: becoming a civil servant or joining a Big Tech company.

That's changing. Especially over the past year, many AI startups have emerged, and more young people are choosing entrepreneurship. AI has created a new path outside Big Tech.

By 2025, not being involved in AI at all will feel like staying in the PC internet era of 2010 instead of joining the mobile internet wave.

Do you have a story to share about working in a Chinese tech company? Contact this reporter at cmlee@insider.com.