Raymond Zhao, 23, is the CEO and co-founder of Structured AI, a Y Combinator Fall 2025 batch startup building an AI copilot for preconstruction engineering. His journey from dreaming of a career at Goldman Sachs to leading his own AI company is a story of redefining success on his own terms. At Oxford, where he studied mathematics and statistics, Zhao initially saw finance as the path to prestige and impact. The idea of working at Goldman Sachs, with its high pay, elite reputation, and smart, driven colleagues, felt like the ultimate goal. He landed a competitive summer 2024 analyst internship in London after a grueling selection process. But the reality didn’t match the dream. Long hours, weekend work, and a lack of personal passion for financial markets made the experience draining. While other interns were deeply engaged in market research, Zhao felt disconnected. After the internship, he took a break, backpacking through Southeast Asia. That time of reflection led him back to Oxford for a master’s degree, where he joined the venture capital society. There, he met his future co-founder, Isabel Greenslade, and began to see a different path. He also joined the university’s AI society, where he and Isabel spent a year refining ideas. They eventually connected with Brandon Smith, whose experience in contracting revealed a key pain point: engineers spend too much time on repetitive, text-heavy tasks during preconstruction. That insight sparked Structured AI. The team built AI agents to automate these manual workflows, helping engineers focus on higher-value work. When Zhao was nearing graduation, he wasn’t job hunting. Instead, he and his co-founders applied to Y Combinator for the summer 2025 batch — and were rejected. In June, during his final week at Oxford, he skipped classes and traveled to San Francisco for an AI startup conference. He hadn’t planned to raise money, but friends encouraged him to pitch. He did, and within days, they secured about $500,000 in pre-seed funding. The speed and ease of raising capital surprised him — a sign, he realized, that the market was hungry for real AI solutions. With the funds, they committed fully to the startup. They refined their product, spoke to dozens of engineers, and built a strong sales pipeline. Their second YC application was successful, and they began the Fall 2025 batch in September. YC has been intense. The team works around the clock, turning their Airbnb living room into a makeshift office with three monitors and fold-out desks. Days start as early as 4 a.m. and stretch into the night, filled with calls, events, and product work. Sleep is limited to about six hours, but Zhao is energized. The goal is Demo Day in early December, when they’ll pitch to a room of top investors. The pressure is high, but so is the potential. For Zhao, the biggest risk isn’t failure — it’s not trying at all. He now sees his current path as a dream, one that came later than he expected, but far more meaningful. “Now, the next one is to build a massive company,” he says. “The biggest risk nowadays is not taking one.”