Science & Technology

Ramp’s Bet on the 100x Employee

by Y Combinator

Share:

📚 Main topics

  • From Paribus to RampEric Glyman and his co-founder built Paribus to automatically claim refunds when purchases dropped in price. After selling it to Capital One, they used what they learned about financial institutions to build Ramp around helping businesses save time and money. 1:02
  • Ramp’s value propositionRamp began by combining financial products with software that helps companies spend less, rather than simply offering rewards that encourage more spending. Its next chapter aims to help businesses grow faster and make the money they do spend go further. 8:20
  • AI inside the companyRamp developed tools including Inspect, an internal coding agent, and Glass, a company-wide AI system connected to employees’ permitted data and workflows. The goal is to make company knowledge and capabilities accessible across teams. 24:24
  • A new way to build softwareAI has enabled employees outside engineering, especially customer experience staff, to create pull requests and fix customer problems much faster. That shortens the path from identifying an issue to delivering a product improvement. 44:39

✨ Key takeaways

  • Agents need the ability to actParibus could identify potential savings but was limited in what it could do. Glyman says useful AI systems need access to operational tools, not just the ability to analyze information and make recommendations. 12:59
  • AI adoption is unevenA small group of employees can become dramatically more productive while other teams lag behind, creating bottlenecks. Ramp paused hiring for a quarter to make AI learning a core part of employees’ work. 21:46
  • Measure outcomes, not just spendingGlyman argues that AI budgets should be judged by the value they produce, rather than by a flat spending limit. Ramp is working to connect AI costs to outcomes such as improved conversions or savings. 27:28
  • Shared knowledge can improve workGlass connects company systems, respects individual access permissions, and lets teams share useful skills and workflows. It can also help surface how decisions are made and where policies need to change. 29:59
  • AI could help large companies move fasterGlyman sees high coordination costs and siloed information as major reasons organizations slow down. He believes AI can help larger companies regain some of the speed and adaptability of startups. 36:49

🧠 Lessons learned

  • Find ideas in everyday problemsParibus grew from a frustrating experience with a flight price drop. Glyman’s lesson is to notice problems in ordinary life and test whether software can solve them. 3:08
  • Listen to how users describe a productParibus founders learned that customers’ explanations to friends revealed both what they valued and what made them skeptical. They used that feedback to improve the referral offer and make the product easier to recommend. 6:47
  • Look for evidence of buildingGlyman recommends finding people who have already made things others want, whether that means building a Minecraft business or excelling at technical competitions. 49:50
  • Ask for help directlyFounders can make progress by clearly describing their biggest problem and asking others for advice. Glyman says that asking for help is a powerful habit at any stage of life. 52:24

🧠 Conclusion and next steps

  • Make AI useful across the organizationRamp is extending its internal tools to customers, including onboarding companies to Glass, while continuing to connect AI usage to practical business outcomes. 27:28
  • Build for what becomes possibleGlyman expects cheaper intelligence to enable new products and specialized businesses, and sees the coming years as an opportunity to rethink how companies operate. 43:35
  • Confront problems and keep movingGlyman’s closing advice is to identify where help is needed, explain the problem, and work through it rather than waiting until everything is figured out. 52:24

Transcript excerpt

0:00 Today on How to Build the Future, I'm sitting down with Eric Glyman, co-founder and CEO of RAMP. RAMP started seven years ago as a corporate card with one promise. Save companies time and money. Today, it's a $60 billion platform that runs the financial operations of tens of [music] thousands of businesses. And increasingly, it does all of that work using agents. >> We're sort of at a pace [music] that I think most people would look at and go, "This is kind of remarkable." This is the story of [music] how Ramp is reshaping the way companies are built

0:31 from the ground up. [music] Thanks for being here, Eric. >> Honestly, Garry, it's great to be back. Thanks for having me. >> Welcome back. Uh you and your co-founder, Kareem, uh you built Paribus and we got to work together YC Summer 15. actually you sold it to Capital One and then you went on to start ramp. Walk us through that journey. What was that like you know deciding to apply to YC uh

🔒 The full, searchable transcript is available with Pro.

🔒 Unlock Premium Features

This is a premium feature. Upgrade to unlock unlimited Q&A, transcripts, mindmaps, and translations.

Questions & Answers

Common questions about this video

What problem did Paribus solve, and how did it work?

Paribus automatically sought refunds when something a user bought online dropped in price. It scanned the user’s email for receipts, checked prices and refund policies, and emailed retailers to request eligible refunds. 1:02

What was the key insight behind Ramp’s original financial products?

Eric and his co-founder saw that conventional financial products often rewarded companies when customers spent more, even when that left customers worse off. Ramp aimed to align its incentives with customers by helping them spend less and keep more money. 8:20

Why did Ramp want to operate closer to the movement of money?

At Paribus, the team could identify potential savings but could only ask companies for refunds and hope they agreed. Ramp’s idea was to move beyond recognizing waste and have the ability to stop unwanted spending at the operational layer. 12:59

Why did Ramp pause hiring for a quarter, and what did it hope to achieve?

Ramp paused hiring partly because recruiting and onboarding take time, but especially to make AI a core competency rather than something employees explored only occasionally. The company wanted to give people the conditions to reach an “aha” moment with AI. 23:19

What is Glass, and how does it help Ramp employees?

Glass connects a company’s core systems, including data warehouses and customer-support tools, so employees can work with a more complete view of the business. It also provides shared skills and access controls, allowing useful workflows and knowledge to be reused across teams. 30:05

How has AI changed who can contribute code at Ramp?

Ramp reported that non-engineers created 7,000 pull requests in one quarter, with many coming from the customer experience team. AI makes it possible for employees close to a customer problem to fix software issues directly instead of waiting through a long ticket and development process. 44:39

How does Eric Glyman recommend finding people who are good at building things?

He recommends looking for evidence that someone has built things people use or pay for, rather than relying mainly on résumés. For example, a teenager who built a business on Minecraft servers has demonstrated an ability to make something others want. 50:20

What advice does Eric give founders about asking for help?

Eric says founders should be willing to state the problem they are facing and ask directly for help. He sees it as a powerful habit: confront the issue, explain your reasoning, and invite others to help work through it. 52:24

🔒 Unlock Premium Features

Access to Chat is a premium feature. Upgrade now to unlock unlimited studying tools.

🔒 Unlock Premium Features

Access to Mindmap is a premium feature. Upgrade now to unlock unlimited studying tools.

🔒 Unlock Premium Features

Access to Translation is a premium feature. Upgrade now to unlock unlimited studying tools.

Key Topics

Get unlimited summaries, Q&A, transcripts and more with Pro

Upgrade to Pro

Suggestions

🔒 Unlock Premium Features

Access to AI Suggestions is a premium feature. Upgrade now to unlock unlimited studying tools.