In the Wuling Mountains on the border of Hubei and Chongqing, a local agricultural enterprise called “Youyoucao E’yu” is quietly emerging as a model for digital transformation in the region. The company, which started with the cultivation and deep processing of Chinese medicinal herbs, has recently attracted the attention of more than a dozen nearby small and medium-sized enterprises (SMEs) thanks to a self-developed AI-based production and sales coordination system. They are no longer just focused on growing herbs; they are now using data to cultivate markets.

“In the past, we relied on experience to decide what to plant and how much. Now, AI can analyze price fluctuations in national medicinal herb markets, the impact of climate, and even predict supply-demand gaps three months out,” said Chen Mingyuan, founder of Youyoucao E’yu, pointing to a real-time data screen in the workshop. This system integrates transaction data from 17 major medicinal herb markets across the country, meteorological satellite cloud images, and logistics cost models. Using machine learning algorithms, it provides planting recommendations for cooperative farmers. Last year, the system helped the company avoid the blind expansion of 300 mu of Coptis chinensis, preventing potential losses of more than 2 million yuan.

The case of Youyoucao E’yu reflects the real challenges and breakthroughs for current SMEs using AI. Unlike large enterprises, which invest tens of millions of yuan in AI, SMEs need tools that are “lightweight, implementable, and effective.” Chen Mingyuan admitted that the system encountered setbacks in early development. “We tried directly using AI platforms from big tech companies, but the data didn’t match, and the scenarios didn’t fit. In the end, we had to form a three-person team ourselves, using open-source models to build a ‘makeshift solution,’ which turned out to be the most effective.”

The core of this “makeshift solution” lies in deep exploration of vertical scenarios. Youyoucao E’yu broke down its AI applications into three modules: first, intelligent planting decisions, which generate plans for fertilization, irrigation, and pest and disease warnings based on historical data and real-time environmental parameters; second, automated quality inspection, which uses computer vision to identify the quality and impurities of medicinal herbs, raising the manual sampling rate from 5% to 100%; and third, production-sales matching, which uses natural language processing to automatically capture pharmaceutical companies’ procurement needs and generate quotations, reducing order response time from two days to two hours.

“AI is not about showing off technology; it’s about solving real problems,” said Li Min, an expert at the Hubei Provincial SME Digital Transformation Service Alliance. He believes that Youyoucao E’yu’s approach offers three insights for the industry: first, data doesn’t have to be large and comprehensive; it only needs to solve specific problems. Second, technology doesn’t have to be cutting-edge; it must be suitable for the company’s business processes to be sustainable. Finally, teams don’t have to be large; the key is to have people who understand the business leading the algorithm design.

Currently, Youyoucao E’yu has broken down its AI system into standardized modules and is offering trial access to neighboring tea and fruit farmers. Chen Mingyuan plans to collaborate with 50 small and medium-sized agricultural enterprises along the Hubei-Chongqing border to build a regional shared agricultural AI platform by the end of 2025. “We don’t need everyone to reinvent the wheel. Together, we can run faster by using the same wheel,” he said.

Under the demonstration effect of Youyoucao E’yu, local governments have also begun to adjust their industrial support policies. A relevant official from the Enshi Prefecture Economic and Information Bureau revealed that a special fund will be established in 2024 to support SMEs in AI application trials in areas such as intelligent quality inspection and supply chain optimization, with third-party service providers offering “AI clinic”-style diagnostics. An experiment in AI implementation, initiated by grassroots enterprises, is quietly spreading across the mountains of Hubei and Chongqing.