The Finance Company That Promised Less Spending

Karim Atiyeh grew up in Lebanon and moved to the United States for university. At Harvard he studied electrical and computer engineering, then computer science. His founder biography is often told through finance, but the earlier discipline is systems: how components interact, where information becomes a decision, and what happens when a process fails under load.

He worked in consulting and learned to code before starting a company. That sequence combined an institutional view of business with the ability to build. It also placed him between two founder types that are usually separated in public narratives: the commercial operator who understands incentives and the engineer who wants the machinery to behave predictably.

Atiyeh co-founded Paribus in 2014 with Eric Glyman and Gene Lee. The service looked for price-adjustment opportunities after a consumer purchase and attempted to recover the difference. Retail policies already allowed the refund; most people lacked the awareness, time, or persistence to claim it.

The company was an early lesson in aligning software with money left on the table. It also taught the founders how fragile automation becomes when it depends on inbox access, merchant policies, identity, and systems outside the product's control. Capital One acquired Paribus in 2016, bringing the team inside a large financial institution and extending Atiyeh and Glyman's working partnership.

Before founding Ramp in 2019, the team interviewed hundreds of business owners and finance professionals. They heard a familiar collection of problems: expense software that created work after a purchase, fragmented tools, limited oversight, and card rewards that encouraged more spending rather than better spending.

The founders chose a provocative position. Ramp would issue the card and participate in transaction economics, but its software would help the customer reduce unnecessary cost. The company launched to its first customer in 2019 and publicly in February 2020. The card was the opening; financial operations became the larger product.

Helping a customer spend less is powerful because it sounds falsifiable. Duplicate software can be canceled. A contract can be renegotiated. A policy can stop an unauthorized purchase. A receipt and accounting code can be captured without employee follow-up. Each improvement can produce money, time, or control.

It is also easy to inflate the category called savings. Cash back, negotiated discounts, avoided purchases, projected opportunities, and employee time are not interchangeable. Ramp's alignment claim remains credible only when customers can inspect the method and distinguish realized value from an estimate. The company should be held to the clarity its own thesis invites.

Eric Glyman served as Ramp's primary chief executive and public narrator through the company's first years. Inside the company, Atiyeh led across risk, operations, marketing, and the systems connecting teams. Glyman's 2026 announcement described a twelve-year partnership and named Atiyeh co-CEO, making the internal division of labor more visible.

This matters because fast-growing financial companies often fail between functions. A sales promise changes risk. A risk rule changes activation. A product automation changes operations and support. Atiyeh's role has been to treat those handoffs as a system rather than a collection of departments optimizing separately.

Ramp expanded into expense management, accounts payable, procurement, travel, treasury, and accounting workflows. The strategic logic is data continuity: the earlier the platform sees an intended purchase and the later it follows the transaction into the books, the more it can automate and control.

The same breadth creates concentration risk for the customer. One vendor can reduce reconciliation and integration work, but an outage, incorrect rule, security failure, or locked account can affect many finance functions at once. Platform convenience increases the standard for resilience, portability, permissions, and support.

By 2026 Ramp was placing artificial intelligence deeper into finance work. Reading invoices, coding expenses, identifying anomalies, answering policy questions, preparing approvals, and negotiating routine purchases all present opportunities for models to reduce repetitive labor.

Finance is also where a plausible error becomes a real payment, a denied employee, a missed close, or a false fraud signal. AI systems need scoped authority, explicit confidence boundaries, audit trails, human review, and reversal. The product must make the reason for an action available after the speed of the action has faded.

In October 2026 Ramp announced a $1.85 billion financing at a $60 billion valuation. It reported 85,000 business customers, a $2 billion annualized revenue figure, positive cash flow, and more than $1 billion in monthly customer savings. TechCrunch reported the financing and scale while placing it inside intense investor demand for AI-enabled software.

These figures establish that Ramp has become a major financial-technology company. They do not settle the quality of its credit, the durability of its valuation, or the precision of every savings claim. Atiyeh's deeper record will be written in the systems that remain after the financing moment: whether customers retain control, whether automation earns trust, and whether a company paid through financial activity can continue to prosper by helping its customers do less of the wrong kind.