China’s AI Stack Is Catching the West

ATP is presented by:

Welcome back to The Asia Tech Podcast. Today we discussed China’s rapidly developing AI stack, the future of intelligent financial infrastructure, AI-powered hiring in hospitality, agentic content operations, shifting AI adoption patterns, and digital identity for autonomous agents.

• David Jenkins - Chief Product and Technology Officer at Openmarkets

• Ayush Soota - Founder and CEO at Paathz

• Jennifer Chen - CEO at Muse AI

• Rachiket Arya - Head of AI at Aspire

• Yifan He - Founder & CEO at Red Date Technology

Here is an overview of the topics we discussed today:

DeepSeek is raising another $7.4 billion — this time at a $74 billion valuation, 50% higher than just weeks ago. What’s driving this for a company that built its reputation on doing AI cheaply? The answer goes far beyond one company. A Shanghai startup called DFSX has also unveiled its own AI chip, the DF100. Built on older 14-nanometer technology, it places memory physically close to the computing layer — reducing the bottleneck that slows AI inference. And DFSX plans to release its next chip, the DF2000, by Q4 2026, claiming it will outperform Nvidia’s H200 — the only chip Nvidia can currently sell in China. If true, what’s left for Nvidia in that market?

DeepSeek’s investors include Tencent, CATL, and China’s National AI Fund. This is not a startup story. It’s an infrastructure story. China is building a complete domestic AI stack — models on one side, chips on the other — much like it approached electric vehicles and semiconductors. Could DeepSeek become China’s answer to OpenAI? And what happens to Western AI companies if China no longer depends on their chips or their models?

David joined us from Openmarkets, one of Australia’s largest execution and clearing brokers. He sees two kinds of investors right now: digital natives pushing hard on multi-agent systems, and traditional investors still rebalancing funds once a year with a financial advisor. These two groups are moving at completely different speeds. Who does AI serve first — and what does that mean for the rest?

David also raised a sharp point about stablecoins. If younger investors in emerging markets buy US assets using dollar stablecoins, why would they ever convert back to local currency? This could quietly drain liquidity from smaller economies. Most regulators haven’t caught up to that risk yet.

Ayush built Paathz specifically for hospitality hiring — and he reframes the whole problem. The industry faces a projected 40 million worker shortfall by 2030. But Ayush argues there are plenty of qualified candidates. The real issue is signal-to-noise. A luxury hotel receptionist and a budget hotel receptionist have completely different skill sets — and a standard CV cannot tell that difference. Traditional job boards are flooded with applicants who have never worked in hospitality. Can AI actually surface the right candidates, or does it just produce a new kind of noise?

Jennifer runs Muse AI, working with global consumer brands on content. The biggest shift she sees: AI is no longer a tool that people operate. It’s becoming a layer that businesses run on. Her team has flipped the traditional marketing workflow — instead of validating content after it runs, they validate before anything is produced. They build digital clones of consumer personas that simulate how target audiences might respond to creative, running 24 hours a day. What could brands learn if they could interview their consumer continuously — not just once in a focus group that’s already out of date by the time results come back?

Rachiket is Head of AI at Aspire, a finance platform for global SMEs. His team analyzed spending across 10,000 businesses. In 2025, companies spent 4.2 times more on OpenAI than Anthropic. In 2026, that ratio has dropped to 1.5. Anthropic’s customer base grew 258% in a year and spending is up 17x. His explanation? MCP — the open protocol that connects Claude to other tools — changed who could use AI. Suddenly non-technical teams could build their own workflows. Rachiket’s team is also working toward a bold goal: 100% of their code written by agents, with no human writing code at all. Are we closer to that than most people realize?

Yifan has been building infrastructure for digital currency and digital identity for five years — and says the things he built were never really designed for humans. They were designed for machines. Now that autonomous agents are handling financial transactions, those systems are finally being used the way they were always meant to be.

His work links legal identity to AI agents using zero-knowledge proofs. An agent can prove it has been verified without revealing who it belongs to. When billions of agents are making billions of requests daily, verification cannot go through a centralized government gateway — it has to be decentralized. And when something goes wrong, someone has to be accountable. Right now, most of the world has not figured out how that works yet.

If these threads interest you — digital identity, agentic finance, China’s AI ambitions, and the race to produce content at scale — the full episode is worth your time.

Watch it here:

Watch the episode

Twilio is the secure and scalable platform to build amazing customer engagement without the complexity, with the AI-driven workflows you need to connect.

Visit Twilio.com to start building today

Boomi, the data activation company for AI, powers the agentic enterprise by bringing data to life across the business. Trusted by over 30,000 customers and supported by a network of 800+ partners, Boomi helps organizations of all sizes achieve agility, efficiency, and innovation at scale.

Visit www.boomi.com to learn more

HCLTech is supercharging progress with AI that moves beyond ambition and delivers real business impact. A recognized AI leader, trusted by global enterprises to turn technology into transformation. HCLTech: Supercharging Progress

Visit https://www.hcltech.com/ai to learn more