Interview: The Round Trip

Compiled & edited by: Yuliya, PANews

As the AI wave sweeps the globe, a new era of the internet, dominated by machines (such as AI agents) rather than humans, is quietly dawning. To enable machines to collaborate smoothly and securely in this new world, a programmable trust infrastructure that unifies identity, payments, and governance becomes crucial. Kite AI was born in this context, dedicated to building the world's first AI-driven payment blockchain network, and has already attracted investment from top firms such as PayPal Ventures, Coinbase Ventures, and General Catalyst.

In the new Founder's Talk series of "The Round Trip," co-produced by PANews and Web3.com Ventures, hosts John Scianna and Cassidy Huang invited Chi Zhang, co-founder and CEO of Kite AI, to share her journey from a top AI and data company to entrepreneurship, Kite AI's grand vision of building a trust framework for the machine internet, and how to seize the "once-in-a-lifetime" opportunity in a new AI paradigm composed of data, computing power, models, and agents.

Building Trust for the Machine Internet: Kite AI's Vision and Mission

Host: Welcome Chi. Could you please introduce yourself and your vision for Kite AI?

Chi Zhang: I have a background in AI and big data. I earned my PhD in Machine Learning and Artificial Intelligence from UC Berkeley. After graduation, I joined DotData, one of the earliest and largest automated machine learning platforms at the time, where I led data science and product-related work serving various vertical industries such as healthcare and finance (e.g., banks using AI for transaction fraud detection, and healthcare using AI for medical image-assisted diagnosis). Later, I joined Databricks, where I was responsible for product management of data engineering solutions. Subsequently, I co-founded Kite with co-founder Scott.

Our mission at Kite is simple: we believe that the future internet will be dominated by machine roles (such as AI agents), and their numbers will even exceed those of humans, and this is becoming a reality.

In order for these machines to seamlessly and smoothly complete various workflows for humans, businesses, and organizations on the internet (such as helping you buy daily necessities or assisting companies with recruitment and interviews), these AI agents must possess three core capabilities:

• Identity: Having identity verification to prove "who they are".

• Payment: Possesses the ability to make and receive payments instantly and securely.

• Staying in control: We want everything they do to be done within clear limits and with clear guidelines to ensure they don't get out of control.

So this is basically the original intention and goal of our founding of Kite: to build a programmable trust infrastructure, or a framework, that unifies identity, payment and governance in a programmable way, thereby allowing AI agents to perform tasks on behalf of humans or any subject under explicit guidance.

Host: In your previous work experience, was there ever a moment of "inspiration" that gave you a unique insight and made you realize, "This is the direction we must move in"?

Chi Zhang: I've always believed that data is one of the four core pillars of AI (in recent years, discussions about the pillars of AI have gradually added the dimension of "Agent" to the existing "data, computing power, and models"). During my PhD studies, I focused on training models and causal inference. Later, I delved into data engineering at Databricks. I personally experienced that data is actually the biggest bottleneck for companies doing AI—especially high-quality, unique, and novel data, which is crucial for training models. Even though many people consider computing power a bottleneck (which is one of the reasons why Nvidia's stock price has skyrocketed in the past two years), I still believe that data is an important and urgent issue, especially when we talk about Agents.

Now, when we talk about AI agents, if you speak with companies focused on agents or data infrastructure, they'll all point out that one of the most urgent and critical bottlenecks to making agents truly effective is solving the underlying data infrastructure problem. For example, if you want an agent to help you with cryptocurrency trading, it needs real-time access to massive amounts of data, such as API data on cryptocurrency prices and sentiment data from various social media platforms like Twitter. This requires the agent to have the ability to access and process unstructured or semi-structured data in real time, which remains a bottleneck in the current data infrastructure.

Returning to your question about when I first saw this unique opportunity and combination, I think it really started in the latter half of last year. That's when we began to see significant improvements in the capabilities of agents, which became especially clear at the beginning of this year. At that time, Manus launched its general-purpose AI agent, OpenAI released the ChatGPT Operator, and many other companies demonstrated tremendous progress in the workflow, accuracy, and intelligence of their agents, requiring very little human intervention.

All of this has led me and my team to this realization: if we're not going to become a computing power company like GPU cloud services, or a business like Scale AI or Databricks (although I still think they are among the greatest companies in the industry), then what is the next truly huge opportunity for a startup like ours to focus on?

I believe the answer lies in AI agents or agent-based infrastructure. This is arguably a once-in-a-lifetime opportunity. If you look back at the development of the internet, from 1995 to now, it has been over 30 years, and it has primarily been built for humans—from desktop internet to mobile internet, many designs, such as identity verification, the need to enter CVV codes when paying with credit cards, and the aesthetic design of website front-ends, are all aimed at optimizing human vision, hearing, and the overall experience.

However, I believe that the form of the Internet will be completely different for future "machine actors" or AI agents.

Kite AI's Position in the Proxy Payment Ecosystem

Host: It seems that "agent payments" have become a major focus for many people, with companies like Visa, Google, and the x402 protocol jointly released by Coinbase earlier this year all making moves in this area. So, does Kite AI have a clear direction in mind? What are your thoughts on payments between agents?

Chi Zhang: We warmly welcome and endorse the emergence of these open standards because they are public goods that contribute to industry development. Simply put, the system we built is 100% compatible with these open standards (like x402, A2A, AP2), which are crucial for agent-related transactions and payments.

Our focus is more on the underlying infrastructure: building the foundation for the settlement and verification layer, allowing all agent-related payment transactions or identity verification operations to be completed on this layer. You can think of it this way: protocols like x402, A2A, and AP2 are more like different token standards such as ERC-20 and ERC-721 in the Ethereum ecosystem. And we, Kite AI, are like the Ethereum blockchain itself, the underlying platform built to execute agent transactions (including payments) under these "standards".

We are excited about the emergence of these standards and believe that public goods driven by large companies usually require large companies to promote them, rather than startups alone.

Strategic Financing and Market Strategies: The Value Behind $33 Million

Host: You recently completed a $33 million funding round, with investors including PayPal Ventures, General Catalyst, and Coinbase Ventures, which John just mentioned. We're curious how this funding will specifically help you accelerate your ambitious product roadmap?

Chi Zhang: If you look at our equity structure, you'll see that most of the investors we choose are for strategic partnerships, and many are corporate venture capital firms. Even giant financial VCs like General Catalyst are known for their hands-on approach to their portfolio companies. So, almost every investor we bring in has a strategic purpose behind it.

Some can provide us with distribution channels, or have the potential to provide such channels; some can bring us key personal connections to help us connect with partners such as agent service providers; and some have regional coverage capabilities, such as Japan's SBI Group, which can effectively help us achieve growth in the Japanese market.

So what I want to say is that while funding is important, the connections and resources behind the funding, as well as the value that investors on our equity structure can bring, are the more interesting and exciting parts for us.

This brings us back to the question of how to accelerate our roadmap. I believe that any infrastructure company wanting to drive the adoption and use of its technology must start with a few key verticals. You need to focus your efforts on one to three sectors where you believe the applications will take off the fastest. This is why our investor composition is so significant. For example, PayPal is a prime example of online business payments, especially e-commerce payments. Therefore, e-commerce agents are also one of our key investment areas. We believe that a huge opportunity in the agent payment scenario is to help users (whether individual consumers or businesses) complete their purchases through agents. For individuals, this might mean buying daily necessities, booking flights and hotels; for businesses, it's more like corporate procurement, such as purchasing office supplies, or helping a car company procure auto parts and components globally.

Therefore, this is more of a market entry strategy for us. The funding from investors will primarily be used to help us execute this strategy, including recruiting top AI and other professionals in Silicon Valley and elsewhere.

Host: Since you have so many investors with corporate backgrounds, how will they influence your product roadmap? For example, will they tell you directly what their needs are? Is this also one of the reasons that prompted you to decide to work with Brevis? Because I understand that you will be using their ZK (zero-knowledge proof) technology to develop "ZK Passport".

Chi Zhang: We have a deep connection with Brevis. In fact, we knew Michael (the founder of Brevis) before Brevis and Kite came along. At that time, he was still working on Celer Network, and I was still working on other projects. We were friends. My co-founder's graduate school classmate was also his classmate, so we've known each other for a long time.

The core of payments lies in trust, and the foundation of trust is identity . This concept was inspired by discussions with industry giants like PayPal and Visa. PayPal, as the world's largest personal identity network, has successfully validated the importance of identity in the payment system. Kite aims to create a programmable trust infrastructure that integrates identity, payments, and governance to form a "trust layer" supporting the entire payment system . In this process, ZK technology becomes a crucial tool for achieving identity verification and privacy protection. Brevis's solution provides key support at the verification level and also offers a state channel-like processing solution for high-frequency trading scenarios.

From the early stages of the project, the Kite team recognized that the Agent's transaction or payment needs would be processed at machine speed , not human speed. This transaction model requires ultra-high frequency, real-time, and high throughput, but current mainstream blockchain technologies, including Solana and Ethereum, cannot meet these requirements. To achieve real-time, low-cost, intensive, high-frequency trading, the Kite team explored state channels as a solution. During this process, Brevis technology came into the team's view. Its technical characteristics highly aligned with Kite's core needs, providing the possibility of promoting an efficient transaction model.

The unique challenges of building scalable architectures

Host: It sounds like you're building a well-thought-out solution. Do you currently have design partners helping you shape this system? Or are the ideas you just mentioned already industry consensus?

Chi Zhang: Actually, I'd like to answer from two perspectives. First, we do have some design partners involved, some of whom are on our investor list and some are not, but they have all brought very exciting application scenarios and collaboration opportunities, which we are actively pursuing.

Second, regarding whether this is a "consensus" or "common sense," I would say "yes and no."

The "yes" part is that many of the relevant technologies and concepts already exist.

But what I mean by "no" is that you need a very unique combination of perspectives and a deep understanding of multiple fields such as blockchain, agents, and payments to truly see how to correctly combine these elements. That's why I believe that currently not many people have truly built or designed a truly suitable architecture to solve this problem.

To give another example, if you talk to someone who works on infrastructure or architecture, they'll tell you that in the early stages of a project, there might be thousands of ways to architect a system to solve the same problem. But as the system grows and traffic surges, those thousands of methods quickly dwindle to perhaps only 10, or even 5, viable paths.

We invest a lot of thought, effort, and practical testing in order to build from the very beginning one of the "5 methods" that will ultimately succeed, rather than one of the "1000 methods" that will crumble as the system expands.