He Sold A Company For $50 Million, Runs An Incubator For Aspiring Founders, And Has Built A Generative AI Math Tutor and Teaching Assistant for Higher Ed

It’s not every day you meet someone who was born without a name and went on to become an early backer of one of the most iconic communication platforms of our time—Discord. This sentence just touches on Peter Relan’s journey, which has culminated in impressive outcomes.

Peter transitioned from a refugee settlement in India to building and investing in billion-dollar companies, which is more than a rags-to-riches tale; it impresses with its resilience, clarity, and founder-market fit.

Listen to the full podcast episode and review the transcript here.

A Life That Almost Wasn’t: Peter’s Humble Origins

Peter’s story begins in one of the most unexpected places. He was born at home in a refugee camp in Delhi, India, to parents who had fled Pakistan as teenagers. His birth certificate read simply: “A boy has been born.” No name. No identity.

Just the beginning of a story that would span continents and industries. Peter’s parents, a street cleaner and a seamstress, lived in poverty for years but prioritized one thing above all: education. Peter was the fourth child.

Thanks to their perseverance, he attended a Catholic school where he was given the name Peter. It was there he fell in love with Math, a passion that would shape the rest of his life.

From Delhi to UCLA: A Leap of Faith

Peter’s brilliance earned him admission to IIT, one of India’s most competitive institutions, which usually starts in July. But he had bigger dreams. Simultaneously, he applied to the University of California, which begins in October, and got accepted thanks to his impressive SAT scores.

Thus, Peter chose to pursue his undergraduate studies at UCLA instead, a bold move that most Indian students typically made at the graduate student level. Despite having limited financial resources, he managed to make it work, with some help from his brother and by tutoring Math to earn a living.

That move, as Peter puts it, was a calculated risk to enter the “land of opportunity” as early as possible, before his peers at IIT. He was also interested in STEM and technology, and he knew that Silicon Valley and California were where he ultimately wanted to be.

The Corporate Training Ground: HP, Oracle, and Stanford

Surrounded by the likes of Mark Benioff and inspired by Stanford’s legendary entrepreneurship classes, which he attended while pursuing his M.S. through an HP Fellowship, Peter felt the startup itch, but knew he wasn’t ready just yet.

He recalls listening to notable guest speakers, including Scott McNealy, the founder of Sun Microsystems, and other prominent figures from the software industry. However, Peter was aware that he needed not just a strong launchpad and a risk-oriented approach, but the right timing to build startups.

After completing his graduate studies at Stanford in 1992, Peter joined Oracle and became VP of the Internet division. This was during the early days of the Internet and the Netscape boom, and he would work here for the next four years.

Later, Peter handed the position over to Thomas Kurian, another Oracle colleague, now Google Cloud CEO.

Joining Webvan as the CTO

Although Peter had several ideas for startups, he wanted the experience of working in one. Instead of jumping headfirst into founding his own company, he joined Webvan as a founding team member and CTO. Louis Borders, also the founder of Borders Books and Music, created Webvan.

Webvan is a hyper-growth startup backed by Benchmark and Sequoia that went from Series A to IPO in under two years. As Peter sees it, the company was the easiest launch pad, as it had venture capitalists managing the executive recruitment process.

Webvan built one of the world’s largest automated distribution centers, and Peter’s job was to make the software that ran it. He recalls that he didn’t fully understand the business model, but that wasn’t part of his job description, per se.

The company went public as one of the biggest IPOs ever in Silicon Valley.

The Webvan Lesson: Tech Alone Isn’t Enough

Webvan is simultaneously considered a pioneer of its time and a huge success, yet also regarded as a poster child of the dot-com bust. At the time of its IPO, it was valued at $8B, and $12B at its peak. Netscape went public at $6B.

But Webvan’s downfall taught Peter a critical lesson: no matter how brilliant the tech, a risky business model can sink the ship. Ultimately, Kaiser Permanente acquired Webvan’s technology. Kaiser is a giant automated distribution center and software to deliver hospital supplies and medication.

Peter knew that he had built amazing technology and was obsessed with the idea that there must be some way to make technology successful as a business. This experience planted the seed for his next chapter, one in which business fundamentals would be just as critical as technical innovation.

Building Business Signatures–on 9/11

In a remarkable display of determination, Peter incorporated his next venture, Business Signatures, just before September 11, 2001. The world was in chaos, and he was fully aware that the dot-com bust would result in a nuclear freeze in the venture capital market for at least two to three years,

But Peter saw two possibilities: either everything would fall apart, or those who built through adversity would emerge stronger. More business would be conducted online. But the 9/11 terrorist attack meant that the world wasn’t going to support the company for a while.

However, Peter was determined to continue. He wanted to become “one of the most persevering and diligent people to make it through one of the worst downturns in the 21st century.”

Business Signatures was founded with two co-founders from Peter’s Oracle days and Dave Roux, ex-Oracle and founder of the tech buyout firm Silver Lake, as an angel investor.

It focused on a pressing problem of the internet era: online fraud. Now that the Internet and eCommerce were big things, people were concerned about their credit cards and bank data being stolen.

People were reluctant to use their credit cards online, and banking was gaining momentum. Peter and his team developed technology that could analyze people’s click streams in real time, enabling businesses to operate on the Internet and monitor transactions.

The software would identify signatures of fraudulent behaviors and was thus called Business Signatures. The company sold enterprise software to banks that wanted to bring online banking to their customers and provide them with a fraud detection system based on early rules-based AI

Business Signatures served clients including Citibank, U.S. Bank, and even USAA. Entrust eventually acquired the company for over $50M, validating Peter’s vision of combining robust technology with business acumen.

An Unexpected Acquisition Offer

Talking about the acquisition, Peter reveals how they had just raised venture capital from TPG Ventures and Walden International, two well-known firms. Business Signatures was approached by a public company that was trying to crack the same space from a different angle.

At the time, Peter and his co-founders were in the process of raising their series B round. Initially, the discussion with Entrust’s CEO was about financing Business Signatures. However, as the business dinner progressed, it transitioned into an acquisition offer.

Storytelling is everything that Peter Relan was able to master. The key is capturing the essence of what you are doing in 15 to 20 slides. For a winning deck, take a look at the pitch deck template created by Silicon Valley legend, Peter Thiel (see it here), where the most critical slides are highlighted.

Remember to unlock the pitch deck template that founders worldwide are using to raise millions below.

Around the same time, Peter’s main competitors, Passmark and Cota, were acquired by RSA, a security company. The competitive environment had suddenly grown rapidly, and Peter began to consider scaling their software with the help of a larger company like Entrust

Entrust had the sales team and infrastructure, and it made sense to partner with someone with a distribution channel to assist in scaling. Peter accepted the offer.

Founding YouWeb: Scaling Entrepreneurial Impact

Having experienced the highs and lows of building a startup, Peter asked himself a big question: “How can I scale my impact beyond one company at a time?” He wanted to put some of the fortune from the acquisition to work, not as a venture capitalist but as an entrepreneur.

The answer was YouWeb, an incubator designed to back technologist-founders with deep founder-market fit. Peter envisioned being a collaborator, joining the board, helping with fundraising, and strategizing their technology. He also wanted to offer mentoring and access to his network and expertise.

But unlike accelerators like YC, YouWeb provided a year-long salary to a “founder-in-residence” to explore and validate their ideas. It was co-creation at its finest. Over the past 18 years, YouWeb has incubated 35+ companies, including:

• ThreatMetrix (acquired for $830M by LexisNexis)
• OpenFeint (acquired for $104M)
• CrowdStar (acquired for $50M)
• Discord (last valuation at $15B)

The YouWeb Filter: People Over Ideas

Peter talks about his criteria and framework for evaluating startups. He answers the question to, “How do you know which idea is worth exploring, worth investing in, or someone worth investing in versus something to filter quickly and move on to the next one?”

• Technologist First – Builders with the capability to create.
• Founder-Market Fit – Deep understanding or passion for the domain.
• Clarity + Focus – Visionaries who know what they will do and what to say no to.

Peter explains that these traits made Jason Citron, the founder of both OpenFeint and Discord, a repeat success. “Passion can be misdirected. Founders who know what not to do are just as important as those who know what to build.”

Peter also cites the example of two Australian founders who had provided Business Signatures data on IP address and geolocation.

Post the Entrust acquisition, they were keenly interested in continuing with their version of a fraud detection business. They focused on the geolocation of IP addresses and the botnet detection market. That company was to become ThreatMetrix

That was an example of Peter’s belief in backing founders with ideas based on their understanding of the customer pain points. They had the founder-market fit because they had studied or been in that industry.

Peter has also encountered many founders who are reluctant to acknowledge or address the issues they may face. They are unwilling to think outside the box about what can make the company successful. Or, take risks to change or pivot. He believes passion is good, but flexibility is crucial.

The Power of Saying No

One of the most critical traits Peter looks for in founders is the ability to say no. In an age of distractions and shiny objects, staying focused on the core mission is a superpower. At YouWeb, Peter encourages founders to walk away from ideas that don’t show traction, even after a year of funding.

This disciplined approach has led to a 20% hit rate, consistent with industry benchmarks, but with far more intentionality and minimal burn. Peter reveals that YouWeb has been in operation for 18 years and has helped build around two new companies every year, bringing the total to 35 to 40 so far.

A Full-Circle Journey: Championing Education with MathGPT.ai

Peter has returned to his original inspiration as his career matures: education and learning. Having gone from an unnamed child in a refugee settlement to a Stanford-educated founder and investor, he knows firsthand the transformative power of learning.

Peter’s project? MathGPT.ai is a Gen AI Math Teaching Assistant and conversational Tutor transforming how instructors and students approach one of the most fundamental yet intimidating subjects in learning.

Peter’s passion for math increased during his undergrad years at UCLA, where he took a job tutoring math and STEM subjects to support himself.

But what struck Peter wasn’t just the subject but the students. He watched freshmen and sophomores make major life decisions, such as changing majors or career paths, because they couldn’t get through math, algebra, or calculus midterms.

Whether aspiring to be doctors, nurses, psychologists, or even sports journalists, non-engineering students were abandoning their dreams simply because they couldn’t overcome math hurdles. Close to 90% of non-hardcore STEM majors still needed a math course. That experience left a lasting impression.

Decades later, when a founder approached Peter with the idea for a conversational AI-assisted Math tutoring network, he immediately saw the transformative potential.

Building the Foundation: Real Conversations, Real Data

Before AI could fully step in, the team spent years building a network of AI-assisted human tutors who offered real-time Math and STEM support and conversational tutoring sessions via text messages for underprivileged students across the country.

Partnering with over 100 colleges across the U.S. who distributed their product, they facilitated over 5 million STEAM tutoring sessions, gathering invaluable insights into student struggles and behavior. This human-first approach wasn’t just a stopgap but a strategic move.

It generated a massive dataset that would later become the moat for training generative AI models. Peter explains that MathGPT.ai has developed and delivered a full AI Teaching Assistant and Tutor to colleges today, in higher education. It caters to hundreds of classrooms and is instructor-led.

Building MathGPT.ai: A Scalable, Ethical AI TA & Math Tutor

With the arrival of ChatGPT in late 2022, Peter and his team launched MathGPT.ai in January 2023. The product wasn’t designed to replace teachers; it was built to empower them. MathGPT.ai offers:

• Instructor-guided course and assignment management
• Guardrails to prevent cheating and make LLMs extremely accurate
• AI-driven tutoring support tailored to college-level coursework

The platform has delivered nearly 200,000 AI-led Math tutoring sessions and is used by instructors at hundreds of colleges. It’s scalable, accessible, and most importantly, ethically aligned for academic integrity.

As Peter explains, the instructor assigns reading material and tasks, and the tutor is available to assist students.

MathGPT is a personal fulfillment for Peter to see the same college students that he tutored 40 years ago as a human being, but on a limited scale.

The Vision: Universal Math Confidence

Peter sees a future where every student from 10th grade onward has access to Math tutoring that removes fear and builds confidence. He wants every student to use MathGPT.ai through instructor-led, and ultimately, self-paced or tutor-led learning.

The platform offers both a free version, where instructors can introduce AI assistance to students, and a paid version, which includes course management features for institutional use, running the classroom under supervision.

“Math is everywhere,” Peter explains. “From buying insurance to taking out a mortgage, understanding math means understanding life. We don’t want students to say, ‘I can’t do this.’ We want them to say, ‘I have a choice, I choose this or that.’”

That choice to pursue STEM, to navigate life’s financial decisions, to make purchases, or simply to understand the world better should never be lost to math anxiety. Even as he champions AI, Peter is clear about its limits.

“Artificial intelligence brings intelligence, but humans bring wisdom,” Peter says. That’s why MathGPT is designed to be instructor-led or tutor-led, because the fusion of human insight with AI scalability is where real educational change happens.

A Final Reflection: Advice to a Younger Founder

When asked what advice he’d give to a younger version of himself, Peter is quick to respond: “Know what you don’t know. You may not understand it yet, but stay open to learning.” Balancing confidence with humility is a recurring theme in Peter’s story.

He says, “It’s very difficult to thread the needle between the confidence of an entrepreneur who says–I know what I’m doing–and simultaneously have a personality of a little bit of paranoia and a little bit of–I don’t really know what I don’t know.”

Peter believes that self-awareness, paired with relentless curiosity, creates founders who can not only survive but thrive through the unpredictable startup journey.

In Conclusion

From refugee camps to AI-powered classrooms, Peter Relan’s journey proves that education and learning, especially Math education, can be a gateway to personal and societal transformation.

With MathGPT.ai, he’s not just building a product; he’s creating a transformative experience. He’s building a world where no student has to abandon their potential because they “just aren’t good at Math.”

Whether it’s mentoring founders, investing in edtech, or simply passing on his playbook, Peter Relan’s career has come full circle, proof that where you start doesn’t determine where you’ll finish.

Listen to the full podcast episode to know more, including:

• Founder-market fit and technical ability are non-negotiables for startup success.
• Clarity of focus—knowing what not to do—is as critical as passion.
• Great technology without a sound business model is not a recipe for success.
• Entrepreneurship requires both confidence and the humility to keep learning.
• AI’s true power lies in scaling impact, but human wisdom must guide it.
• Universal math confidence can unlock careers and decision-making for millions.
• Incubators like YouWeb prove that empowering builders early leads to outsized outcomes.

 

SUBSCRIBE ON:

For a winning deck, see the commentary on a pitch deck from an Uber competitor that has raised over $400M (see it here). 

Remember to unlock for free the pitch deck template that founders worldwide are using to raise millions below.

 

Podcast: Play in new window | Download

Subscribe: Apple Podcasts | Spotify | TuneIn | RSS | More