According to Business Insider, Dan Popescu, Goldman Sachs’ head of AI engineering for asset management who was just promoted to managing director among 638 employees last week, revealed the three essential skills for AI roles at the bank. He identified AI engineering knowledge, finance expertise, and traditional software engineering as the ideal combination needed to succeed. Popescu, who earned his PhD in applied mathematics and statistics from Johns Hopkins University in 2021, leads a team creating AI tools for portfolio managers and advisers that integrate proprietary data to enhance workflows. The 36-year-old executive works directly with investors in the asset management group, combining the firm’s data with AI capabilities to maximize speed and efficiency. His promotion signals that managing directors at Goldman aren’t all dealmakers – some focus on creating tech tools that could ultimately boost the bank’s bottom line.
Wall Street’s rare breed
Here’s the thing about Popescu’s trifecta of skills – it’s actually pretty rare to find all three in one person. Most tech professionals either lean toward pure engineering or they understand finance but can’t code. The real value comes from sitting at that intersection where you can speak both languages fluently. Popescu calls his team the “middleman” using a “wheel-and-spoke model” to connect core engineering with business-facing developers. That’s basically the sweet spot where AI actually delivers value rather than just being shiny new technology.
Goldman’s AI ambition
This isn’t just about one team or one division – Popescu mentions the “One Goldman Sachs” initiative that aims to break down barriers between teams using AI across the entire firm. CEO David Solomon recently said he wishes the firm’s $6 billion technology budget were even higher, and he expects headcount to trend upward thanks to AI efficiencies. So we’re talking about a massive, firm-wide transformation here. The bank is using its own AI-powered copilots to speed up software development and holding internal trainings to spread these capabilities. It’s a classic case of eating your own cooking, but on a billion-dollar scale.
Wall Street’s AI arms race
Goldman isn’t alone in this push – JPMorgan, Morgan Stanley, and other rivals are all racing to integrate AI across their operations. Startups like Hebbia are trying to capitalize on Wall Street’s AI craze too. But bringing a 150-year-old institution like Goldman into the AI age represents a multi-billion dollar effort that will shape the industry’s direction for years. The real challenge? Finding enough people who actually have that magic combination of skills Popescu described. When you need someone who can both build sophisticated AI models and understand why a particular trading strategy matters, you’re fishing in a pretty small pond.
The PhD advantage
Popescu’s background in applied mathematics and statistics gives him an interesting perspective on what it takes to succeed in this environment. He says his doctoral research forced him to think about the same academic problem for five years, which developed the persistence needed for complex AI projects. “Persevere and develop that grit and strength to continue working on something even if you don’t immediately see the light of day,” he advises. That’s actually pretty crucial when you’re dealing with AI implementations that might take years to show real returns. The patience required for academic research translates surprisingly well to the marathon of transforming a financial giant’s technology stack.
