From Berkeley to Uber China, to Echelon AI: Exploration and New Beginnings with Eddie Guo

Mother of Success (MOS): Hi Eddie, could you start by introducing yourself to our readers?

Eddie Guo: I’m Eddie Guo, born and raised in China. I came to the U.S. in 2011 for college and have been working in tech ever since. I started my career with Uber’s China Growth team during an incredibly high-growth period, and later I joined an early-stage AI startup. In 2023, I started my first company - and it’s hard to believe it’s already been nearly two years. Time really flies.

College Years at Berkeley: Exploration and Integration

MOS: What was your college experience like at Berkeley?

Eddie: Two words: exploration and integration.

Academically, I switched majors four times. I started in Environmental Science because I’ve always loved nature - in high school, I even wrote my college essay about the relationship between humans and the environment. But after a year, I found it too broad - it touched policy, biology, chemistry, physics, everything. I took classes in environmental policy and economics but couldn’t picture what I’d actually do with it.

So I switched to Statistics. Then to Operations Research, which blended math, programming, and analytics in a more applied way. Later, I added Computer Science after realizing how much I enjoyed building things. Eventually, I graduated with a double major in Operations Research and Computer Science.

Outside academics, I joined a fraternity - a hundred-year-old red-brick house that appeared in The Graduate. People often associate fraternities with parties, but for me it was about community and tradition. The walls were lined with portraits of alumni who went on to become doctors, lawyers, even Antarctic explorers. Living there for over two years, I learned about the American tradition of charity and built friendships that have lasted to this day.

After Graduation: Why Uber China

MOS: With a CS degree, you had many options. Why Uber China?

Eddie: I’ve always wanted to start a company, but for my first job, I wanted to be where innovation was happening - and at the time, that was the sharing economy.

Even in college, I had tried small experiments in entrepreneurship. A few friends and I were trying to make and sell alcoholic jello shots - it didn’t go far, but it gave me a taste of creating something from scratch.

Uber China was unique because it was one of the few Silicon Valley companies operating deeply in the Chinese consumer market. The team was expanding into new cities every week, and our group worked on subsidy optimization - figuring out how to compete effectively with local rivals using data and incentives.

When Uber China merged with Didi in 2016, it marked the end of an intense but formative chapter. That experience taught me how large-scale systems grow - and how startup energy can thrive inside a fast-moving company.

Transition to B2B SaaS: A New Learning Curve

MOS: Later, you joined a B2B startup. What was that like?

Eddie: In 2019, I joined an AI startup building internal IT tools for enterprises. It was around 60–70 people when I joined - still early-stage but already working with large customers.

The biggest lessons I took away were:

• B2B and B2C are completely different.
  B2C scales through network effects; B2B grows through trust, relationships, and word-of-mouth.

• Product development is more instinctive.
  At Uber, every decision required A/B tests. In B2B, you can’t always wait that long - you build, show it to customers, learn, and iterate quickly based on their direct feedback.

• Business is human.
  Even in enterprise software, decisions are made by people. Building trust often matters more than adding features - something technical founders tend to underestimate at first.

The First Startup: AI On-Call

MOS: When you started your own company, what was your initial idea?

Eddie: Our first idea was to build AI-powered on-call engineers - using AI to help developers handle alerts and incidents.

The idea came from personal pain. At my previous company, engineers were constantly woken up at night by system alerts - hundreds every week - which killed daytime productivity. On-call management was a huge problem.

Three of us decided to quit and build a solution. It was scary - no customers, no funding, no product. But we felt the timing was right. The AI wave was just beginning, and life felt too short not to try.

Lessons from the First Company

MOS: You worked on that company for over a year. Why did it eventually stop?

Eddie: Two main reasons:

• Demand wasn’t strong enough.
  We did long proof-of-concepts with large engineering teams. While managers liked it, executives didn’t view it as essential enough to buy - it solved pain, but not a must-pay problem.
  

• To solve the problem well, it requires so much access, and getting that access takes a long time.
  
  We realized that doing it right meant deep integration with each company’s systems, tools, and workflows. That level of access is hard to obtain early on, especially in enterprise environments where data sensitivity and permissions take months to navigate. 

The lessons were humbling but invaluable:

• Talk about budget early - free pilots rarely convert.

• Validate demand quickly.

• Domain expertise isn’t enough; you need a painkiller, not a vitamin.

Pivot and Fast Validation

MOS: How did you pivot afterward?

Eddie: After we sunset the first product, we gave ourselves one month for each idea we identified - 50 customer calls, no excuses. If we didn’t find conviction, we’d move on to the next idea.

That period taught me that passion alone isn’t enough - startups have to solve painful, valuable problems. We were transparent with our team about the reset, and everyone appreciated the honesty.

We started from first principles:

• Identify our strengths.

• Find large, AI-disruptable markets.

• Match our skills to the right space.
  

We realized our real strengths were in product and B2B SaaS, not developer infrastructure. I started researching enterprise IT services and the ServiceNow ecosystem - reading industry reports, studying M&A data, and talking to dozens of people across the value chain.

Market Validation and Echelon’s Turning Point

MOS: How do you decide if a market is big enough?

Eddie: I usually look for at least a $1B total addressable market. In enterprise SaaS, $100M ARR is a real business - if you can capture even 10%, you have something meaningful.

MOS: You attended the ServiceNow conference in Vegas this year, right?

Eddie: Yes, and that was a turning point.

By then, we’d already done over 100 interviews across the ServiceNow ecosystem - customers, partners, salespeople, even training agencies. We decided to test demand by creating momentum: launch demo videos, open booking links, and start taking meetings.

At the three-day conference, we met 20–30 people - mostly at the Starbucks near the venue. It was nonstop demos and conversations. Those in-person discussions moved faster than six months of online outreach. Some of those early believers became our first customers - and even friends we still grab beers with today. That trip gave us our first real conviction that we were on to something.

What’s Next for Echelon AI

MOS: What’s next for Echelon?

Eddie: Since May, several customers have gone live. We’re now expanding deployments and increasing product intelligence.

Our goal is to build a fully automated enterprise AI agent that assists with development, testing, and documentation - effectively augmenting how enterprise IT teams work. Enterprise IT service is a trillion-dollar market, and as foundation models evolve, more workflows can be automated end-to-end. That’s the opportunity we’re chasing.

Final Thoughts

MOS: Lastly, could you recommend a film or podcast to our readers?

Eddie: Big Fish (2003). It reminds me that life itself is an adventure - you have to embrace the unknown and imagination. Even the unexpected parts eventually become part of your story.

MOS: That’s wonderful. Thank you for sharing, Eddie!