- Yan Zhang merges AI and crypto with a personal approach, building human-like digital agents through his project Delysium.
- Beyond Delysium, Zhang explores AI-blockchain intersections with other stealth projects and advisory roles across global tech ventures.
While most people are busy finishing their thesis, Yan Zhang is already immersed in a noisier world—the sound of fans from his self-built mining rig. He may have graduated from Beihang University and earned a master’s degree from Carnegie Mellon, but the interesting story is not about his degree, but his decision to mine crypto using GPUs while still working at NVIDIA.
He knew exactly how the chips worked, so why not take advantage of it? While coding for the autonomous vehicle system, he also mined ETH in his spare time. Not just random multitasking.
But the turning point did not come because the hash rate or Bitcoin price soared. Zhang felt that AI and blockchain technology had a meeting point that was more than just hype. That was the beginning of RCT AI, and not long after that—Delysium was born.
Building Delysium: When AI and Crypto Are No Longer Far Apart
Delysium is not an ordinary Web3 project built just for the sake of having a token. This platform offers something a bit strange but intriguing: AI-Twins. Imagine creating a “digital twin” that you can chat with, and even help you manage your crypto portfolio. Sounds like a typical AI application? Unfortunately, it’s not that simple.
Delysium’s AI-Twins aren’t just chatbots. They have “personalities” and can move around in the metaverse ecosystem, not just sit on a chat screen. Zhang said that AI should be an active agent, not a passive one—and to do that, it needs infrastructure that supports scalability and decentralization. This is where Delysium’s role as an AI agent framework really comes into play.
The platform is even connected to the Ethereum and BNB Chain networks via the “Lucy” feature, a kind of AI OS that can be used to search for on-chain data, execute transaction commands, and even check token liquidity—all through plain text commands.
Yan Zhang: Blending AI and Crypto With a Personal Touch
On the other hand, Delysium’s steps have also accelerated after receiving funding from various major investors, including Galaxy Interactive and Republic Crypto. It doesn’t stop there, Google Cloud and Microsoft have also formed strategic partnerships to support its AI side.
It’s safe to say that this isn’t an indie project that’s just going ahead. But strangely, it still feels like a project born from the garage of an idealistic programmer.
Zhang himself is still active behind the scenes. He is also involved in other projects such as Deterrence and Soularis. Some of these projects have not been widely discussed, but from his style and approach, it seems that he is not finished exploring the boundaries between AI and blockchain. He even became an advisor to Point72 Ventures, highlighting AI projects outside of Asia.
However, what is most interesting about Yan Zhang is not only his achievements, but the way he bridges advanced technology with concepts that feel personal. Rather than creating a cold and overly technical AI system, he invites us to imagine a more human future—although ironically, through digital agents.
Furthermore, his approach also shows one thing: the world of AI and crypto are no longer two different worlds. They absorb each other. Zhang seems to be a translator in the middle of the two—with GPUs, code, and a little curiosity that never dies.
Sometimes, the stories of successful people in the crypto world sound like fairy tales full of numbers and graphs. But in Yan Zhang’s case, what is interesting is the “digging” side. From tinkering with mining rigs, building AI systems for cars, to creating AI-Twins that can discuss tokens, it all started with a simple logic: if technology can help humans become smarter, why should it be limited?
Maybe not everyone is ready to chat with their digital twin. But if you look at the direction Yan Zhang is pointing, the crypto world in the future seems to be more crowded—and more eccentric—than we think.
