AI Transforms Fourth-Tier City Rehabilitation Center: 40% Profit Growth

Introduction: When AI Enters the Most "Human-Dependent" Industry

In July 2026, at Sunflower Rehabilitation Center in Pingdingshan, Henan Province, a young therapist was conducting social training with an autistic child. On the screen before her, the RICE AI system developed by Dami & Xiaomi—a leading institution in special needs children's rehabilitation in China—was synchronously recording the child's behavioral data, analyzing intervention progress, and generating course recommendations for the next phase.

This is not a future scenario but an industry transformation already underway. In third- and fourth-tier cities like Pingdingshan, Cangnan, and Lianyungang, this system has been operating for over six months.

Industry Dilemma: Million-Person Talent Gap, Scaling Bottleneck Locked

The special needs children's rehabilitation industry has long faced a structural problem: not enough people.

According to Ministry of Education public data, there were 912,000 special education students nationwide in 2023, with only 77,000 full-time teachers. When including more types of special needs children such as autism and language development delays, the industry generally estimates actual demand at the tens of millions level. By professional配比, the practitioner shortage is counted in millions.

The bottleneck exists not only in quantity but also in quality. Zeng Songtian, Executive Director of the Dami & Xiaomi Children and Youth Development Institute, explained that the core shortcoming in China's special needs rehabilitation industry is the lack of structured precipitation and data accumulation of intervention methods and programs for children's common and individualized problems. Frontline therapists need to conduct classes under supervisor guidance, and cultivating a qualified supervisor takes several years, with only a handful of supervisors nationwide with rich practical experience.

In third- and fourth-tier cities, talent scarcity is even more pronounced. Song Wentao, head of Pingdingshan Sunflower in Henan, said there are very few local special education rehabilitation talents, and senior therapists and supervisors are "difficult to recruit even with high salaries." To get their children systematic rehabilitation, parents have to travel back and forth between first- and second-tier cities, spending at least 100,000 yuan per year. Many families give up halfway, and children miss the optimal intervention window.

Even in first- and second-tier cities, therapists' working conditions are not optimistic. Nong Xinyi, head of Nanning Nuan Yang Children's Rehabilitation Center in Guangxi, reported that therapists spend大量 time on paperwork, reports, and teaching aid production. New employees have long training cycles, and labor costs gradually increase without being noticed.

This is not the predicament of individual institutions but a structural bottleneck for the entire industry. When service supply completely depends on people's time and experience, the ceiling for规模化 is locked at the speed of talent cultivation and retention rate.

RICE AI: AI-Assisted Therapists, Improving Children's Core Social Disorders

In October 2024, Dami & Xiaomi launched RICE AI—a generative AI system for the entire special education rehabilitation process.

Wang Zhanwei, R&D Vice President of Dami & Xiaomi and CEO of Zhongmi Internet, described RICE AI's positioning this way: "AI cannot replace therapists. The value of AI is to make the rehabilitation process from assessment to intervention simpler, allowing more people to do this work, reducing rehabilitation costs, and enabling more children to receive good rehabilitation services."

Specifically, RICE AI intervenes in three环节: assisting in completing ability assessments and automatically generating reports, reducing completion time from hours to minutes; outputting personalized intervention plans based on massive cases, executed after supervisor review and confirmation; and providing decision references for young therapists approaching the level of senior supervisors.

Zeng Songtian explained that the core of special needs children's rehabilitation is real interaction between people, which AI cannot replace. However, AI can standardize processes and digitize experience so practitioners don't have to start from scratch every time. "The value of therapist-child interaction is irreplaceable, but AI can amplify therapists' professional capabilities."

Clinical Data Validation: Significant Improvements in Five Key Areas

Zeng Songtian cited data demonstrating changes in special needs children after institutions used RICE AI.

"This is a before-and-after comparison using PEP-3 (standard assessment tool) conducted by three third-party institutions on two groups totaling 65 cases (average age 4.3 years). Analysis showed that the experimental group (using RICE AI) achieved significant intervention results in five key areas: social interaction, language expression and communication, non-verbal expression and communication, cognitive performance, and imitation ability."

Specifically, the experimental group showed the most significant improvement in "non-verbal expression and communication," with post-test scores increasing by 5.3 points compared to pre-test. In "language expression and communication," the experimental group also showed a significant growth trend, with average scores increasing by 5.2 points, significantly outperforming the control group's growth rate.

In cognitive and imitation abilities, data also supported the effectiveness of the intervention program. The experimental group's scores in "cognitive performance" increased from 78.0 to 83.5. In "social interaction," the experimental group also showed significantly better progress than the control group, further confirming the positive effect of RICE AI-assisted intervention programs on improving children's core social disorders.

Business Results: 10x Faster Assessment, 40% Profit Growth

The effectiveness of RICE AI courses for children was reflected in changes in parents' choices.

Pingdingshan Sunflower was one of the first institutions to introduce RICE AI. Song Wentao provided straightforward comparative data: previously, from a child's entry assessment to plan proposal, it took an average of two to three days; now it can be completed in two to three hours, with efficiency improved by over 10 times. Parent waiting times have decreased.

More important changes appeared at the commercial level. Sunflower launched the "RICE Social Training Course" based on RICE AI, and one parent purchased 1,000 classes at once. In the first half of 2026, the institution's profit increased by 40% year-on-year. Song Wentao attributed the growth to expanded production capacity from improved service efficiency, as well as increased parent renewal willingness driven by quantifiable results.

Wenzhou Cangnan County Fangzhou Rehabilitation Center showed similar trends. Chairman Chen Yu'ou said, "After introducing RICE AI, traditional 30-minute courses switched to 45-minute standard class hours, and the employment model is also changing—previously, parents only recognized senior teachers, but now with standardized teaching materials provided by the system, young therapists get up to speed quickly, and classroom quality is more stable. The institution is no longer constrained by senior teachers, and employment concerns are completely resolved. Just one week after the system launched, we added more than ten new students."

When Wenzhou Disabled Persons' Federation leaders visited the institution for research, they specifically learned about RICE AI's application model and highly recognized this AI rehabilitation system. Chen Yu'ou stated that AI special education application special training will be conducted later to let this model benefit more special needs children in Wenzhou.

Feedback from Lianyungang Xinkang Baby Integration Kindergarten in Jiangsu also pointed to cost reduction and efficiency improvement. Director Huang Pengyi introduced that RICE AI provides full-process standardized new employee training courses, helping new teachers quickly establish professional awareness and master basic teaching procedures. New teachers with relevant experience can teach independently in half a month, significantly shortening the training cycle.

More importantly, the standardized system provided by RICE AI unified teaching standards across all staff. The institution no longer faces被动 situations from core teacher departures, and management pressure is reduced. After the first batch of children using RICE AI received services for one and a half months, parent satisfaction reached 100%.

Industry Transformation: From Direct-Operation Institutions to Industry Infrastructure

In 2014, Jiang Yingshuang, a former journalist, founded Dami & Xiaomi because her daughter was diagnosed with "suspected autism." Over 12 years, the company opened more than 80 direct-operation centers in over 30 cities nationwide.

Enabling more children to enjoy high-quality rehabilitation services has been Dami & Xiaomi's most fundamental belief since its founding and the starting point for all explorations. Based on this belief, Dami & Xiaomi adheres to two foundations: first, insisting on scientific evidence-based practice; second, believing that technology will eventually transform this industry even when AI still seemed "clumsy," hence long-term deep cultivation offline to prepare for the future.

In 2025, RICE AI officially opened to the public, shifting from serving its own institutions to serving the entire industry.

As of now, over 300 institutions have completed registration, with more than 100 using it regularly. Users cover provinces and regions including Henan, Hubei, Jiangsu, Zhejiang, Guangxi, Hainan, Guangdong, Shandong, and Xinjiang, with implementations from first-tier cities to border areas.

This path took 12 years. RICE AI's technical barriers come from 12 years of accumulating over 40,000 real cases, 200 million intervention data points, and 3 million supervisor decision messages. Wang Zhanwei explained that general large models lack professional corpus and contextual understanding in the special education industry and cannot generate truly usable intervention plans. RICE AI performs model fine-tuning based on massive high-quality vertical big data accumulated over 12 years, outputting personalized plans that fit industry standards.

In terms of product form, RICE AI covers the entire process: assessment, planning, lesson preparation, teaching, feedback, and goal adjustment. Therapists are liberated from tedious paperwork and teaching aid production, focusing their energy on interactive teaching with children—the most irreplaceable part of special needs rehabilitation.

"We are transforming from a chain intervention institution to industry infrastructure. On top of the industry's infrastructure, every institution—including Dami & Xiaomi's direct stores and ecological partners—can develop their own characteristics," Wang Zhanwei revealed that RICE AI continues to advance. As the demand for school-age special needs children continues to grow (in addition to autism, there are diverse issues such as learning disabilities and ADHD), Dami & Xiaomi has been continuously expanding its service lifecycle in recent years, creating an exclusive school-age version system.

In August 2026, the RICE AI school-age version will launch, focusing on campus integration, classroom adaptation, and academic support. Over 100 schools have applied (including special schools, regular schools, and kindergartens), with strong market demand.

"Let AI do the technical part, let therapists return to the most essential companionship and guidance, and achieve rehabilitation inclusiveness through technology." Wang Zhanwei stated that this is what RICE AI has been doing and the direction Dami & Xiaomi firmly follows.

Conclusion: A Model for AI Empowering Traditional Service Industries

The successful application of RICE AI in the special needs children's rehabilitation field provides a valuable model for how AI can empower traditional service industries. In this most "human-dependent" industry, AI is not a replacement but an enabler—it amplifies professionals' capabilities, lowers service thresholds, improves service quality, and ultimately achieves a win-win situation for both commercial and social value.

From first-tier city scenarios sinking to third- and fourth-tier cities, from self-owned institutions expanding to the entire industry, RICE AI's story continues. For more traditional service industries facing talent bottlenecks and efficiency bottlenecks, this may be a transformation path worth learning from.

Source: QbitAI · July 20, 2026