从售前到亿万富翁

AI technology is transforming every industry at an unprecedented pace. Among the most compelling stories is that of Elias Torres, former VP of Engineering at HubSpot and founder of Drift. This young man immigrated to the United States from Nicaragua at age 17 without speaking a word of English—yet through sheer intellect and determination, he wrote a real-life "American Dream" story in the tech world.

Torres's career is nothing short of legendary. During his tenure at HubSpot, he built deep expertise in customer management. He then founded Drift and successfully sold the company for $1.2 billion (approximately RMB 8.7 billion), once again proving his business acumen. Yet his entrepreneurial journey didn't stop there. About a year ago, Torres launched Agency, a startup dedicated to revolutionizing customer success management with AI technology. The company quickly secured $12 million in seed funding, demonstrating the massive market demand for AI-powered customer management solutions.

So how did Torres come up with the idea of using AI to manage customers? In a recent episode of the podcast No Priors, he sat down with host Sarah to share his entrepreneurial journey and his unique insights on AI-empowered customer success. Torres explained that he deeply recognized the limitations of traditional customer management approaches and decided to use AI to break through those constraints. Through AI, he aims to understand customer needs more efficiently, predict customer behavior, and deliver more personalized service experiences.

The following content is based on the podcast video, with some additions and deletions made by InfoQ.

Key takeaways:

• Future companies should empower customers with control, delivering personalized, customer-centric service experiences even when serving hundreds of thousands of clients.

• As companies scale, the need for business growth becomes more urgent—this is when you must solve the "how to scale effectively" problem.

• Building the product is the hardest part, because large language models only solve a small piece of the puzzle. We still need to build products and solutions that address specific problems.

• While technology can help us handle certain tasks, true trust and adoption still take time.

• For the things you're not good at, let them be.

Sarah: Please introduce yourself briefly.

Elias: I'm from Nicaragua, a first-generation immigrant who initially spoke no English at all. After spending ten years at IBM, I decided I wanted to be an entrepreneur. But inside IBM, I couldn't find the freedom I needed—I needed the opportunity to create a bigger impact. Eventually, I moved to Florida and have been here ever since.

Sarah: I feel like the tech world has noticed a pattern: many smart kids who end up in tech participated in math competitions growing up. Were you one of those kids?

Elias: Yes. In Nicaragua, I was selected to represent my school in national competitions and ultimately placed third nationwide. It wasn't a monumental achievement, but the key point is that math was critical along the way—mathematical thinking has real practical value.

Later, I came to the United States and lived in a low-income neighborhood in Tampa, Florida. I still couldn't speak English. But once again, a math teacher asked if I wanted to join the math competition team. I said yes and joined the "nerd club." I struggled with word problems, but I did pretty well on everything else.

Sarah: You've held leadership roles at companies of various sizes, especially after your success at HubSpot where you led the engineering team. Why did you still decide to found Drift?

Elias: I feel like my whole life has been a journey of exploring the unknown, never having a clear goal or foreseeing what would happen next. Like when I competed in math competitions in Nicaragua—I got to represent my country in Mexico and meet kids from all over Latin America. That experience was incredible. Then I came to the U.S., ended up in Florida, and started getting exposed to places like MIT. So I've always felt that each step was like taking on a new challenge, with more opportunities waiting for me. The whole process has been fascinating.

When I was at HubSpot, the company had 200 employees and only $30 million in revenue. Brian Halligan, HubSpot's co-founder and executive chairman, told me: "Our goal is to take the company public within a year or two." At the time, I had no idea what going public even meant. But everything went smoothly over the next three years—HubSpot successfully IPO'd, becoming a $1 billion company, with revenue growing from $30 million to $130 million.

So I thought: if Brian could do it, why couldn't I? I was really naive back then—I thought going from $0 to $100 million would be easy. That's ultimately why I left HubSpot.

Sarah: The company you founded, Drift, eventually sold for over $1 billion. Yet despite that, you still view it as a failure in some respects. Why do you describe it that way?

Elias: Honestly, I'm very satisfied with my life. I barely had food as a kid, so for me, every year has been better than the last. But when we left HubSpot and decided to start our own company, I left it at a $1 billion valuation and watched it later grow to $3 billion. So during the journey from founding Drift to selling it, I realized I hadn't fully achieved the ambitions I originally set for myself.

The day we sold Drift was incredibly sad for me. I felt like a failure. That night around 11 PM, I was lying in bed at home, completely exhausted. Then I got a text from a lawyer saying the deal was done, and I felt nothing. The next day, it felt even stranger and more surreal. It wasn't until much later that I began to understand what selling the company and getting $1 billion in cash really meant. We spent eight years taking the company from $0 to $100 million (approximately RMB 720 million) in revenue. We put in enormous effort, but I also lost the dream of building a $3 billion company or taking it public.

At that point, I was truly exhausted. I had been working continuously for 45 years without ever really taking a vacation. I started to doubt whether I could keep going. If you had asked me then whether I'd start a new project, I probably would have said: "I'm done. I don't want to start anything again." If someone told me, "I want to start a company," I would have thought it was foolish.

Sarah: So you took some time off, and then you started consulting for OpenAI, helping their customers. Where did that motivation come from?

Elias: It was November 2022, right when ChatGPT launched. I happened to be in Brazil—actually, it was my first real vacation after the acquisition. I have a hobby: kite surfing. One morning, I got up and went kite surfing with three friends, enjoying the ocean breeze. Then we sat down for breakfast, and I started explaining ChatGPT to them.