American financial markets move in milliseconds, but the infrastructure underneath them still depends on traders picking up phones, brokers brokering deals by hand, and platforms running on technology that has not meaningfully modernized in decades.
Sangjun Yum, a 34-year-old senior software engineer at crypto prime brokerage FalconX, has spent the better part of a decade watching that contradiction play out up close. From structured product marketplaces to institutional crypto trading, he has seen the same digitization gap repeat itself at every level of the market, and at FalconX, he’s working to close it.
Realizing the Limitations of the Financial System Early On
Sangjun Yum did not decide to become a software engineer in a computer science classroom. The decision took shape around 2008, when he was a teenager watching the American housing crisis shape up and trying to make sense of what had gone wrong.
His fixation was specific and, partly, universal: how could anyone have seen it coming? The financiers clearly hadn’t. Yum’s answer pointed elsewhere.
“I began thinking it can’t be only the finance people being shocked by this,” he recalls thinking. “There must be some kind of engineer who can actually come up with some kind of tooling to predict the crash.” The idea that navigating a system as complex as the financial markets required computational tools, not just financial intuition, became the lens through which he would approach his entire career. It also planted an early skepticism about Wall Street’s relationship with technology, one that would be confirmed at nearly every stop along the way.
The 2008 crash wasn’t the only thing shaping his thinking during those years. By the time he reached college, Yum had already identified himself as a conservative investor, drawn to structured notes and market-linked CDs: instruments designed for capital protection and steady, defined returns, typically used by retirees and high-net-worth individuals. He wanted to invest in them, but without a registered financial advisor, the products were simply out of reach.
That frustration added a second dimension to what had started as an intellectual interest: the digitization of finance wasn’t purely an engineering problem. It was also a question of who got access and who didn’t.
Dealing with the Gap Firsthand
After graduating with a Bachelor’s in Computer Science with a Minor in Mathematics from NYU, Yum began working as an engineer for SIMON Markets, a Goldman Sachs spin-off building a centralized digital marketplace for structured investment products typically available only through registered financial advisors. He was the seventh engineer on the backend team.
The problem SIMON was trying to solve was straightforward in concept: instead of a financial advisor having to sit across from a salesperson to review a prospectus before buying one of these products, why not create a single platform where issuers could upload their products and advisors could access them all in one place?
In practice, the problem was significantly harder. Structured products have no universal data standard. Every issuer structures its products differently, with different life cycles, different terms, and different prospectus formats, meaning there’s no common language. Yum’s job was to build a unified portfolio management tool that could serve both sides of the market: the sell-side issuers uploading their products, and the buy-side advisor networks accessing them.
Satisfying dozens of issuers with competing requirements sharpened his instinct for building flexible, scalable architecture in a way few assignments could have. “Wall Street is not as techie as it sounds,” Yum explains, “and the U.S. financial market is a lot more outdated than I thought.”
The SIMON experience also clarified something about why retail investors remain locked out of structured products in the first place. Technology is part of the answer, but not all of it. By law, these products are restricted to certified registered investment advisors, meaning regulatory gatekeeping is the primary barrier. Technological obsolescence compounds that barrier, but it doesn’t replace it.
It also confirmed something he would carry into every role that followed: the digitization gap in finance was not limited to smaller or less resourced institutions; it was an industry-wide condition.
Modernizing the Infrastructure Behind Crypto
After spending the better part of five years working inside the structured products world, Yum joined FalconX in January 2024, drawn by what he saw as a genuinely unsolved version of the same problem he had been chasing since 2019.
FalconX is a crypto prime brokerage serving institutional clients, from hedge funds to asset managers and large retail platforms, looking to move significant volumes of digital assets without the price slippage that comes from trading directly on retail exchanges like Coinbase. The core problem Yum joined to help solve is the elimination of manual trader intervention from that process: replacing the phone-call-and-handshake model with automated, optimized execution across a network of liquidity providers.
“A lot of trading is done via white-glove, high-touch service,” Yum explains. “I was interested in working more on digitizing those kinds of experiences for the public in general.”
The comparison he draws to make the problem legible is a useful one: just as Uber replaced the dispatcher and DoorDash replaced the phone call to the restaurant, FalconX is working to replace the trader who picks up the phone when a hedge fund wants to move a large position. Each of those earlier industries had a moment where the human intermediary became optional, and Yum believes institutional trading is approaching the same inflection point.
The work itself isn’t consumer-facing. There are no interfaces to design or features a retail user will ever see. Yum is, instead, building the invisible execution layer that determines whether a large institutional order gets filled at the right price, in other words, the infrastructure underneath the infrastructure.
The Invisible Systems Behind Every Crypto Trade
Through his work at FalconX, Yum has developed an unusually clear view of the hidden market structure that sits between a retail investor tapping a button and a trade actually executing.
When a retail platform receives a buy order, it doesn’t route that order directly to an exchange like Coinbase or Binance. Instead, it maintains its own internal books, tracking how much crypto it owes its users and how much exposure it’s carrying at any given moment, and periodically hedges that net exposure through over-the-counter trading networks. Routing large volumes directly to a retail exchange would result in poor pricing. The OTC network absorbs those orders instead, re-hedging them across a network of liquidity providers.
“When you hit the buy or sell button for Bitcoin,” Yum explains, “a large portion of crypto exposures are actually hedged out in the OTC trading network. This is something not that many people realize.”
FalconX operates as one of the primary counterparties in that process, sitting between the retail platforms and the liquidity providers, handling order flow that most users never know exists. Manual processes at this scale introduce latency, pricing inefficiency, and human error into markets that move in milliseconds.
That invisibility is itself part of the problem; a financial system that most people can’t see is one that most people can’t push to improve. Digitization and transparency, in his view, are not separate goals. They tend to arrive together, and after nearly a decade of building inside the machinery most investors never think about, Sangjun Yum is still working on making sure they do.
For him, his work is the most direct expression yet of the thesis he first formed as a teenager in 2008: that finance’s most persistent inefficiencies are, at their core, engineering problems waiting to be solved.