When you join the UC Berkeley Haas MFE program, you'll learn everything you need to break into traditional quant finance. But for those with bigger ambitions, we're the foundation for something more. Chandler Fang MFE17 initially built his career on the proven tracks of Wall Street, proving his mettle as a Quant Researcher at JPMorgan and establishing a prestigious foundation in traditional finance. Yet, his trajectory evolved alongside the financial landscape: leading him through senior product roles at Ripple to his current position as Co-Founder and CEO of t54 labs, where he is now pioneering the trust layer for the agentic economy to allow AI agents to safely transact capital.
More Than a Career
Like many MFE graduates, I started my career as a quant at JPMorgan. I built core skills in statistics, programming, and risk management, and developed a deep understanding of how financial systems operate. But traditional quant roles operate at a remove from everyday economic activity. Over time, I wanted more exposure to the business side and to work closer to real-world applications — which led me to transfer internally into J.P. Morgan's AI and blockchain initiative (Onyx), where I worked across data science and product roles. It was there that technical depth crossed business judgment. Building something from scratch began to feel not just possible, but necessary.
Founding t54
The inflection point was 2023, when tools like ChatGPT made clear that AI systems were evolving from assistants into autonomous actors. The gap became difficult to ignore: today's financial infrastructure is entirely human-centric — identity, authorization, and risk frameworks all assume a human decision-maker. Once agents can take actions, it becomes inevitable that they'll need to interact with financial systems. Blockchain and stablecoin infrastructure made this even more compelling: they're API-native and programmable, which aligns naturally with how agents operate, and they allow non-human actors to hold and transfer value without relying on traditional custodial identity models. t54 came out of this realization: if agents are going to execute economic activity, we need a trust layer designed specifically for them.
Building with the Best
As we started building, one thing became obvious: the industry was moving fast on enabling agent transactions, but far less attention was being paid to risk, accountability, and failure handling. That led us to initiate conversations with researchers at Microsoft Research, Google DeepMind, and Columbia University. The collaboration came together organically where researchers were interested because t54 could provide unique real-world transaction data, while we benefited from their theoretical frameworks and modeling approaches. Together, we co-authored our first paper on quantifying trust in agent financial activity, and we continue to convert that research into real products.
We've also built that trust layer directly into the infrastructure agents are starting to transact on. t54 built and maintains the x402-xrpl package, the reference implementation of x402, an open protocol for HTTP-native machine-to-machine payments, for the XRP Ledger, now featured directly in Ripple's own developer documentation as the way agents pay for X402-gated services using XRP or RLUSD. And in June 2026, t54 joined Mastercard's Agent Pay for Machines as one of its initial participants, bringing its trust layer to this new class of payments and helping make agent-led payments more secure, auditable, and accountable, alongside partners like Coinbase, Stripe, and RippleX.
Read about Chandler's $5 million seed round here.
If you're considering the UC Berkeley Haas MFE, follow us on Instagram and LinkedIn for alumni stories, community updates, and more.
