How a Property Company Built Its First AI Team, Burning Hundreds of Thousands of Tokens Daily

This company has turned all its employees into AI enthusiasts.

On DingTalk, there’s a company like no other: operating in a traditional industry, yet boasting over 100,000 "AI enthusiasts" among its workforce; consuming more than 100 million tokens in a single day, consistently ranking among the top enterprise users on the platform—second only to tech giants and internet companies such as Ant Group and Taobao Flash Purchase.

It is Country Garden Services, China's largest property management company.

While most traditional enterprises are still at the stage of “leaders announcing AI initiatives in meetings, with isolated pilot projects,” this property management leader, managing over 8,000 projects nationwide, has already made AI an integral part of daily work for every employee. Country Garden Services isn’t just “using AI”—it has reshaped itself into a true “AI organization.”

How Did Country Garden Services Become the Leading “AI Organization” in Property Management?

The property management industry is inherently difficult to standardize: housing demands differ between northern and southern regions, facility management varies between old and new buildings, and owner expectations diverge sharply between basic and premium communities. Even within projects managed by the same company, it's hard to achieve uniform service standards and consistent management quality.

Zhao Xiaoguang, CTO of Country Garden Services, mentioned in an exclusive interview with Biji Xian that one of the core challenges in the property sector is the lack of standardization—a hurdle every property company must overcome.

"Although the industry has basic service standards, service experience itself has no upper limit," Zhao said. "Once homeowners get used to the convenience of hailing rides or ordering meals online, they can no longer accept inefficient models—like having to visit a physical service center or wait hours for a response. The principle is exactly the same."

More critically, most property fees go toward labor costs. Security guards, cleaners, concierges, maintenance technicians—all rely heavily on human input. But people only have so much energy and time—24 hours a day. Even if everyone works their hardest, there remains a ceiling to efficiency.

Zhao repeatedly emphasized one term: the technology gap. The more traditional the industry, the greater the potential for efficiency gains.

On one side are rising homeowner expectations for service quality; on the other, ever-increasing labor costs. The widening chasm between them is precisely this "technology gap."

It is against this backdrop that while most companies remain cautious or conduct only small-scale experiments with AI transformation, Country Garden Services has shifted AI from being a "nice-to-have novelty tool" to becoming the underlying way of working across the entire organization. One set of data clearly illustrates this shift: on the entire DingTalk platform, Country Garden Services now consumes over 100 million AI tokens per day. At Country Garden Services, thousands of meeting minutes are automatically generated by AI each day, and over 90% of AI resource usage comes directly from employees—including headquarters staff, regional personnel, project managers, concierges, and customer service representatives.

Country Garden Services does not see itself merely as a "residential community manager," but rather positions itself as a lifestyle service provider. Its central mission has always been how to use technology to improve both service and management efficiency, while also making residents’ lives more comfortable and hassle-free.

Driven by this goal, the company has placed its hopes for breakthroughs on AI, taking the first bold step across the industry's "technology gap" and courageously becoming a pioneer in the field.

Many corporate transformations stall on the same issue: tools designed by headquarters often aren't adopted by frontline staff. The reason is simple—the tools created at headquarters simply don't work well on the ground, adding workload instead of reducing it.

But Country Garden Services took the opposite path.

As Zhang Ruimin, founder of Haier Group, once said: "Successful corporate transformation always combines top-down strategy with bottom-up vitality."

The AI transformation at Country Garden Services wasn’t a one-way push, but a gradually spreading practical process.

The first core tool that was successfully implemented and widely adopted across the company wasn’t developed by headquarters with a million-dollar budget—it was spontaneously built by frontline employees using DingTalk’s AI-powered spreadsheets.

Why did it gain traction? Simply because it actually worked.

Zhao Xiaoguang shared with us: "After DingTalk launched its AI spreadsheet feature, we kept experimenting to see what real problems it could solve. Before we even applied it to inspection workflows, we were already using it for scheduling, tracking special projects, and even replacing many tasks that previously required dedicated software. Gradually, we found the sweet spot where tools meet business needs."

"Meanwhile, we cultivated digital teams in each region who understand both the tools and the business operations. Together with DingTalk, we conducted multiple training sessions to teach frontline teams how to use these tools to build their own processes. This inspection system wasn’t built by IT staff at headquarters—it emerged step by step as frontline teams, after training, realized they could solve problems themselves using the tool."

What Pain Points Did This Frontline-Built AI Tool Solve?

In the past, the process for frontline inspections—from identifying to resolving issues—looked like this:

For example, upon noticing a broken hallway light, staff had to take a photo, return to the office to fill out a form, seek supervisor approval, and then assign the task to an electrician. Ten minutes was considered fast. If any link failed, the issue would easily disappear without resolution, leading to resident complaints—and frontline staff would end up taking the blame.

Now, the process has been dramatically streamlined:

Upon discovering an issue, staff simply pull out their phone, snap a photo, and say into DingTalk: “The hallway light in Building 3, Unit 2 is broken and needs repair.” AI automatically recognizes this as a maintenance request and instantly assigns the work order to the electrician responsible for that area—all completed within about a minute.

After repairs, the technician uploads a photo of the fixed light. The AI can then automatically compare before-and-after images to verify whether the repair was properly done and provide audit suggestions to the project manager.

The long-standing issues of “no response after reporting” and “inability to track progress and outcomes in real time” have been significantly improved, enabling faster responses and continuous follow-ups—truly achieving the goal of “small issues resolved overnight, major ones followed through.”

Beyond this, management can use DingTalk dashboards to monitor key metrics in real time—such as number of work orders, response speed, closure rates, and recurring issues—enabling data-driven optimization of services.

The entire workflow—from reporting, assignment, execution, verification, to improvement—is now fully tracked. Whether problems are resolved well or not is now clear and transparent.

Zhao Xiaoguang added: "Previously, our model involved researching regional needs, extracting commonalities, and building standardized systems—but individualized needs often fell through the cracks. Now, our logic is reversed: frontline staff can build their own solutions using available tools. We then identify successful practices across teams, extract universal capabilities, and integrate them into our core systems.

This creates a new model: instead of headquarters pushing tools down to frontline teams, everyone in the company now actively seeks better tools and embraces AI. Just as non-coders today can use low-code platforms to build websites, the barrier to entry has essentially disappeared."

Clearly, the company’s open attitude toward AI has fostered spontaneous, bottom-up adoption. Headquarters’ ability to quickly capture, refine, and scale excellent frontline practices transforms individual innovations into organizational capabilities—eventually turning them into the company’s “muscle memory,” strengthening service capacity over time.

This bottom-up organizational vitality also solves one of the biggest headaches in large enterprises: distorted communication.

You’ve likely experienced this: a directive clearly defined at headquarters becomes diluted when passed down to regional offices, and by the time it reaches project sites, it’s completely unrecognizable.

The reverse is also true: persistent complaints from residents at the front lines get filtered out layer by layer as they move upward, leaving headquarters unaware of real conditions.

As the book *Complexity* points out: the larger an organization grows, the complexity of its management increases exponentially.

With over 8,000 projects across China, and a multi-layered structure spanning headquarters, regions, local branches, major projects, and individual sites, this problem is especially acute for Country Garden Services.

So how do they tackle it?

They’ve now used DingTalk Meetings and AI transcription to untangle this knot.

On one hand, AI records and summarizes every meeting, making it immediately visible whether meetings were held and whether key messages were conveyed—eliminating room for shortcuts.

On the other hand, genuine resident concerns discussed in frontline meetings can be directly aggregated to headquarters via AI transcription and DingTalk AI spreadsheets—no more层层 written reports or presentations. Information loss during transmission is effectively eliminated.

In the past year alone, Country Garden Services has accumulated 5.38 million hours of AI meeting transcripts. To put that in perspective, it’s equivalent to saving over 600 years of manual note-taking time. Meeting content is now automatically transcribed, key points extracted, and action items generated—turning what used to be pure time cost into valuable digital assets.

AI Coaching: Empowering Every Employee

If an open organizational culture is the soil for a bottom-up AI revolution, then individual empowerment is the root that sustains its growth.

At Country Garden Services, the value of AI ultimately rests on each employee. After all, only by closing the gaps in professional skills and business understanding—raising everyone to a baseline level—can the organization sustain a steady flow of grassroots innovation.

In Zhao Xiaoguang’s view, property service boils down to three things: serving people (homeowners) well, managing facilities well, and handling matters well.

In the past, digitization in property management focused on “control”—monitoring people to ensure standardized procedures were followed. Today, with AI-driven transformation, the focus shifts to “empowerment”—using tools to handle mundane tasks, freeing up human energy for warmer, more thoughtful service.

This aligns perfectly with management expert Henry Mintzberg’s insight: "The essence of management has never been control, but the liberation of human potential."

Take property concierges as an example.

A single concierge often serves hundreds of households with diverse needs: elderly residents living alone may want regular visits and chats, while young professionals prefer minimal disturbance. They must comfort upset complainants and handle emergencies calmly. These nuances and techniques cannot be mastered by memorizing rules—they require repeated practice.

How was this done before? Through centralized offline training. But with tens of thousands of concierges nationwide, how could everyone be trained simultaneously? New hires, lacking experience, would easily panic when dealing with emotional homeowners—often making situations worse.

Now, they use DingTalk’s AI Coach.

New concierges no longer need to memorize regulations. Instead, they use the AI Coach to simulate real-life service scenarios: how to talk to distressed homeowners, how to care for elderly residents, how to respond to emergencies—practicing repeatedly in realistic simulations so they won’t be overwhelmed when facing real situations.

All Country Garden Services needs to do is input course materials and the best practices of top-performing concierges into the system. Then, AI provides personalized coaching to every concierge across the country.

Here’s a concrete example: a new concierge opens a course titled “Home Visit Care for Elderly Residents.” Immediately, the AI transforms into Grandpa Wang—an elderly man with hearing difficulties who speaks slowly—closely simulating a real home visit.

The concierge must follow standard procedures: checking if the anti-slip mat at the door is secure, whether the hallway light works, if the emergency call button functions. They must also patiently ask the elder about sleep quality, diet, and whether medication is taken on time.

Throughout the simulation, AI monitors in real time: Did you complete all steps? Did you slow down your speech? Did you use respectful terms? Were you impatient?

After the session, AI scores performance and gives feedback: “Ask more open-ended questions instead of yes/no ones,” “Emphasize bathroom slip prevention,” etc.—clear and actionable.

"We hope AI coaching provides ongoing training so employees master communication skills across different residents and scenarios, delivering warm, human-centered service," said Zhao Xiaoguang.

AI’s support extends beyond concierge training—it permeates every aspect of property services:

Even the best concierge can’t reply to messages 24/7. Traditional shift patterns inevitably lead to delays at night or during holidays. AI helps bridge this gap—handling routine tasks like checking regulations or submitting repair requests—freeing concierges from administrative burdens.

Similarly, elevators are specialized equipment. Maintenance used to rely entirely on veteran technicians’ experience—whether repairs were adequate or failures predictable depended largely on personal judgment, making standardization difficult. Now, AI analyzes elevator operation data to detect potential faults, significantly enhancing the diagnostic abilities of average maintenance staff.

Through this continuous empowerment of individuals, Country Garden Services not only raises the baseline capability across the organization but also unlocks each employee’s full potential—fueling endless momentum for bottom-up innovation.

The Ultimate Mission of AI: To Free People and Warm Hearts

Every technological revolution in human history has, at its core, aimed to liberate people.

If the steam engine freed humans from heavy physical labor, electrification restructured industrial operations, and informatization broke down information barriers, then this wave of AI is bringing a profound dual liberation to property management—one of the most traditional sectors in daily life:

On one hand, it frees frontline service workers from tedious, repetitive tasks—filling forms, writing reports, memorizing rules—so they’re no longer tied down by bureaucracy and can focus on delivering truly warm, human-centered service.

On the other hand, it narrows the long-standing efficiency and standardization gap in the industry, finally offering a replicable, actionable path for upgrading a highly fragmented field.

The reason Country Garden Services leads the industry lies not in isolated disruptive innovations, but in sustained, deep integration—internally driven by unwavering transformation vision and courage, externally powered by continuous adoption of DingTalk’s full-stack AI products and capabilities.

Zhao Xiaoguang stated plainly: "DingTalk is the core tool for internal management within our entire digital ecosystem. We treat DingTalk as the primary mobile office gateway. Employees open DingTalk to access all work information and complete full business processes—meetings, document sharing, lightweight app development—all closed-loop within DingTalk."

As Jack Ma once said: "All technology and innovation must ultimately return to life itself, create value for society, and bring beauty to people’s lives—or else they are meaningless."

Good AI is never meant to replace people—it exists to serve them.

Good technology doesn’t make service colder; instead, by freeing up human effort, it makes service more professional and more humane.

And what Country Garden Services and DingTalk are jointly building is exactly this: using technology to preserve the original intent of service, employing more efficient tools to gradually bridge the industry’s technology gap, and ultimately delivering more reliable daily services to every community and better living experiences to every homeowner.

We dedicated to serving clients with professional DingTalk solutions. If you'd like to learn more about DingTalk platform applications, feel free to contact our online customer service or email at This email address is being protected from spambots. You need JavaScript enabled to view it.. With a skilled development and operations team and extensive market experience, we’re ready to deliver expert DingTalk services and solutions tailored to your needs!