What Three Days in China Showed Leaders About AI, Platforms, and Data Infrastructure Sequencing

A week inside China’s digital economy revealed a market where payments, logistics, identity, and data systems work together without friction, and why seeing a system at full speed is much more important than simply reading about one.

Whether you’re an operator sequencing a product launch, an investor sizing up a market entry, or a policymaker choosing what to regulate first and what to leave to competition, seeing how decisions play out in-market changes the quality of the decisions that follow. This is what Caribou’s Live Learning experiences are all about.

Read other blogs in this Live Learning series: What Five Days in the Philippines Taught Leaders About the Future of Digital Financial Services

There’s a moment in any well-designed executive learning trip when boardroom theory runs out. I had read much about China’s digital economy, but I was still not ready for the specific, disorienting strangeness of standing in a Hema store in Shanghai, watching customers scan items on the shelves and then leave while bags of groceries traveled along ceiling rails toward a collection point, where a driver picked them up for delivery to the buyer’s home or officewithin thirty minutes of the customer walking out of the door. In this moment report insights ended, and the real learning began.

Caribou recently led a Live Learning journey through Shanghai and Hangzhou with a senior leadership team from a large African telecom group. They came with specific themes to explore: how to responsibly unlock the data sitting inside their own digital platforms to power AI services; how to define a responsible AI charter before scaling further; and how to safely extend credit and inclusive finance offerings in their home market, about what it takes to evolve from a telco into a full technology company, and about what it actually looks like when a digital ecosystem operates at full speed. Three days in China gave them answers, though not always the ones they expected.

On the first day of our Live Learning trip, local expert Yvonne Yang led the team on a self-directed walk along The Bund, Shanghai’s riverside, framed as a challenge to spot innovation in situ, an immersion far more engaging than sitting through a PowerPoint presentation.

WeChat QR codes were everywhere: on restaurant tables, storefront windows, artworks, power bank stations dispensing charged phones for the equivalent of fifty cents. We visited a government-funded startup incubator that could be accessed by WeChat scan. No receptionist, no sign-in sheet, no membership. Just a desk, fast Wi-Fi, and the city’s ambient assumption that you already had the app.

Luckin Coffee operates without a counter. Every order is placed through WeChat. All customer data is captured. Yvonne explained it simply:

WeChat is the layer nearly everyone touches first, because payment rides on top of an app the city was already using to talk to itself. Alipay runs deeper – credit, lending, identity, healthcare – WeChat is the interface, and every transaction that passes through it feeds Tencent’s data infrastructure. The parallel to Caribou’s Philippines Live Learning trip is direct. There, a covered market in Pasig City, Manila, showed how digital adoption and genuine financial participation can diverge: QR codes worked exactly as designed, but the money was cashed out immediately to pay suppliers who only accepted cash. In Shanghai, the divergence barely exists. The city assumed digital participation years ago and has been compounding on that assumption ever since.

Ant Group, Alipay’s parent company, began as a payments product. That framing is now almost comically inadequate. By the time we were sitting with Alipay’s team in Shanghai, the platform was managing AI-driven healthcare consultations through 300 specialist “doctor twins,” running micro-lending to tens of millions of small businesses (over 53 million, per MYBank’s most recent public reporting), and using satellite imagery to verify farmland activity for farmers with no credit history. Alipay Tap, its NFC payment hardware, launched in mid-2024. By April 2025, less than a year later, it had reached over ten million merchants across more than 400 cities, enabling more than 300 interaction scenarios that run from retail checkout to smart lockers to public services. Alipay has gone far beyond digitizing payments and is building the interaction layer for the physical world.

Coming into the visit, the team assumed that Alipay had grown by simply adding services. As our visit showed, the more nuanced reality is that Alipay grew by making each service a foundation for the next. Payments created the identity layer. Identity-enabled credit scoring. Credit scoring enabled lending. Lending created the data for AI underwriting. Each step was taken only when the previous one could support the addition. The default rate on SME lending is, according to MYBank, below 2%—responsible sequencing in practice.

The strategic question this prompted for the leadership group was a simple one: if you already own the payments rail, what is the next defensible layer above it?

Hema, Alibaba’s flagship retail chain, has been called “a living laboratory” for the fusion of online and offline commerce. Although accurate, that description was completely insufficient preparation for actually standing inside one in 2026. Hema seamlessly integrates physical stores, online ordering, data-driven inventory, and rapid local delivery into one digitally connected shopping experience.

Customers move through the store selecting items—by hand, by phone, or through Hema’s app. After payment, they can then simply walk out. The packing happens behind the scenes: automated fulfillment systems batch and assemble orders continuously, then route them overhead to a collection point where drivers pick them up. Thirty minutes after the customer left, the shopping is at their door.

The executive team assumed the technology was going to be the most interesting part of the site visit, but instead the infrastructure sequencing stood out. Hema works because every upstream enabling layer was built before the retail experience was integrated around it: identity infrastructure (Alipay’s ID-verified accounts, increasingly authenticated via facial recognition at checkout), a payments platform, a logistics network, a data system that prices stock dynamically based on freshness and demand. The store itself was the last thing built, not the first.

Retail here functions as the data-collection interface for a much larger machine. What looked like a supermarket was actually a fulfillment operation that happened to have shelves.

An evening fireside chat brought together investors, ecosystem builders, and startup founders to talk about how Shanghai’s innovation engine actually works. The conversation covered VC portfolio strategy, state-private dynamics, and why foreign companies fail at extraordinary rates without a local partner.

That reframe cuts to something the whole China trip demonstrated repeatedly. The competitive advantage in the cities we visited is not any single company or product. It is the density of the infrastructure (payments rails, logistics networks, cloud platforms, AI models, identity systems) that makes new things possible with less friction.

The leadership team came to China with questions about AI. They left with a different question: what is the infrastructure that would make AI genuinely useful at home, and in what sequence does it need to be built?

China’s scale, its state-private coordination model, and the two decades of infrastructure-building since Alipay’s 2004 launch are not replicable in any straightforward sense. The team knew this before arriving, and the trip reinforced it. 

What is transferable is also more foundational: the sequencing logic. Build the payments rail before it is needed. Create the identity layer before it is required. Design the data infrastructure before the use cases exist. None of these innovations existed independently; each depended on infrastructure built earlier. The leadership team witnessed across the trip the compounding effect of this logic.

Live Learning is designed around the questions leadership teams are already wrestling with. The value is less about seeing impressive companies, it is seeing the conditions that produced them, and returning home with a clearer view of what to build next.

If your team is navigating transformation decisions, explore Live Learning at caribou.global or reach out to Shirley Gilbey (shirley@caribou.global) to start a conversation.