Yifan Zhang on Building AI That Tackles Real Problems

About

Yifan Zhang

Yifan Zhang is the Managing Director at AI2 Incubator, where she helps cultivate the next wave of AI startups. With a background as a serial entrepreneur, Yifan co-founded and led two successful ventures—GymPact, a viral fitness app, and Loftium, a real estate startup that helped renters save for homeownership through short-term rentals. Following Loftium's acquisition, Yifan transitioned into venture capital, bringing her hands-on experience to advise and invest in AI-driven startups. She also collaborates with civic organizations in Seattle to foster AI innovation and revitalize the local community.

"It's not as scary as people think. There are people that do this, that are not any smarter than you - not any better than you. That's probably the biggest mental hurdle that people have to get through: it's not like you have to be some special unicorn snowflake to be a founder. If you have the passion, you're willing to put in the work, and you have some insight, you can do it."

Yifan Zhang

"I think it's really important for a founder to have the ability to wear both hats of pessimism and optimism, and nimbly jump from one side to the other.

One moment, you're pure optimism when you're pitching investors, when you're recruiting your team, when you're really motivating people and selling to customers - you have to have a hundred and fifty percent confidence in yourself, your vision, your team and really move the trajectory forward through that optimism.

And then on the other side, as soon as you're done with that pitch, you have to put on your paranoia hat, and think through: 'What are all the ways that I could have done that better? What are all the ways that we can fail? What are all the things I'm not missing?

Pivoting between pessimism and optimism is what drives people to be able to tackle these really hard industries."

As a builder herself, Yifan Zhang believes anyone with passion and dedication can be a founder. The key, she says, is being willing to leap—even if you're terrified.

"Being a founder is not as scary as people think," Yifan says, reflecting on her journey from Harvard undergrad to serial entrepreneur to venture capitalist. "There are people that do this, that are not any smarter than you—not any better than you. That's probably the biggest mental hurdle that people have to get through: You don't have to be a special unicorn snowflake to be a founder. If you have the passion, you're willing to put in the work, and you have some insight, you can do it."

This ethos—that anyone with the right mindset can create something impactful—has shaped Yifan Zhang's entire career. Now, as Managing Director at AI2 Incubator, Zhang draws on her unique entrepreneurial experience, helping founders tackle real-world problems with artificial intelligence. Her story is one of resilience, rapid learning, and community—a path that led her to understand both the dizzying highs and the gut-wrenching realities of entrepreneurship.

Discovering Tech Through Community

Unlike many of her peers at Harvard, Yifan didn’t arrive with a master plan to break into the tech world. Instead, she found her way into the startup ecosystem through her community—falling in with like-minded people who wanted to create something new. "Discovery through community" became a defining theme of her early years and the companies that she founded.

After graduating, Yifan launched her first company, GymPact, a viral fitness app that incentivized exercise by charging users for skipping workouts and rewarding them for staying active. It was, as Yifan describes it, a "trial by fire". She was 22 years old, didn’t know anything about startups, and was thrown into the deep end of an interesting viral growth moment. Yifan and her team went through Techstars, raising funding from Keith Rabois and Max Levchin, and eventually reached profitability, and as she puts it, "made plenty of mistakes along the way."

“I think part of that definitely shapes me. Every single company, I feel like I'm always trying to get better as a founder. Having that mentality also allows me to share some of my failures, my lessons learned with the founders I work with today."

These formative experiences shaped Yifan's entrepreneurial approach and honed her resilience. From her time building GymPact to her next venture, Loftium—a real estate startup that helped renters save for homeownership—she learned to navigate both explosive growth and sudden downturns. When COVID-19 hit, Loftium saw 90% of its revenue vanish in a week. Yifan and her team adapted, extended their runway, and ultimately secured an acquisition by Fly Homes.

These ups and downs—the adrenaline of rapid growth and the stark challenges of survival—have shaped her perspective as an investor. "I think it gives me deep empathy for how hard certain moments are as a founder," Yifan says. "Sometimes founders have only seen success and don't really understand what it's like in those other moments. I try to bring that nuance and empathy to every advisory and investment relationship I have today."

Lessons for AI Founders

As someone who has made the leap from building startups to investing in them, Yifan’s advice to first-time founders is rooted in both empathy and caution. She understands the allure of tackling the flashiest, most competitive consumer markets—but she warns that those markets are often brutal for new founders.

“For a first-time founder, it’s much harder," she explains. "You’re facing competition from incumbents who build much faster, with data and distribution advantages, and also facing masses of other startups. There’s not a lot of leniency in those spaces."

That’s why, at AI2 Incubator, Yifan has leaned towards legacy or "real-world economy" use cases. "Real-world use cases have a little more leniency for first-time founders to make mistakes, areas like manufacturing compliance or mortgages, where having deep relationships and domain knowledge gives you an advantage." Yifan encourages founders to focus on quality and meeting customer expectations—two things she believes have shifted dramatically in the past decade. "Consumers today have been spoiled," she says. "Our standards for quality are much higher. Apps crashing or slow load times—things that were acceptable ten years ago—are now instant deal-breakers." The core lesson? For AI startups, the complexity of the technology means nothing if it doesn’t meet these elevated expectations.

The Future of AI: Focus on the Boring Problems

While the tech world continues to hype AI as the next big thing, Yifan remains pragmatic. She’s excited by AI’s potential but sees the most promise in the less glamorous, process-driven applications—the ones that can quietly transform legacy industries.

“I’m simultaneously terrified of where AI might be in 10 years and unimpressed by where it is today,” Yifan admits. “That’s why I look for industries and places where we can create real value even with the AI capabilities we have today."

Take mortgage underwriting, for example. Yifan’s investment in Friday Harbor aims to use AI to streamline this painstakingly manual process. "Mortgage underwriting involves matching hundreds of borrower documents to complex guidelines. LLMs (large language models) are uniquely capable of holding all of that information and recognizing patterns far faster than a human underwriter."

Building Community and Serendipity

Community, as it was when she first started, remains central to Yifan's work today. She’s working with Seattle’s Office of Economic Development and the Mayor’s office to create an "AI House"—a physical space for founders, researchers, and investors to meet, share ideas, and spark those serendipitous conversations that often lead to breakthroughs.

“Our goal with AI House is to have a space for large-scale events and meetings—a place where leaders in AI can ambiently run into founders and collaborate," Yifan explains. “Just like my early experiences, it’s about creating an environment where people can discover new ideas through community."

As Yifan sees it, AI’s success hinges on ambitious people—people who are willing to solve hard, often boring problems and share their experiences. Her own journey is proof that with passion, resilience, and the right community, almost anything is possible.

"I’m waiting for the moment—and it’s coming—where the underlying technology is so good and so cheap that we’re ready for the next wave of mass market consumer AI," she says. And when that moment arrives, Yifan Zhang will be there, encouraging new founders to take the leap.

"People who know legacy or real world economy use cases have the relationships and the knowledge to build a good product there. I highly encourage AI founders today to lean into what they're really good at."

Yifan Zhang
close

Submit Help Request

attach_fileAdd up to 5 files

Max 2.0 MB per file. Accepted file types: .csv, .jpg, .jpeg, .pdf, .png, .txt