Eyal
Feldman

It started with a mistake that almost cost Eyal Feldman a client. Not a catastrophic error - just an invoice that got lost in the back-and-forth of a corporate approval chain. Somewhere between the vendor, the finance team, and the relevant approvers, it disappeared. The deal nearly went with it.

That moment crystallized something Feldman had been circling for years: accounts payable was broken in a way that made everyone's job harder but no one's fault in particular. The process was too fragmented, too sequential, too dependent on email threads and tribal knowledge. The people who needed to approve invoices weren't in the same room - or even the same system - as the data they needed to act on.

In 2014, Feldman and his brother Ofer co-founded Stampli in Mountain View, California. The founding premise was disarmingly simple: turn an invoice into a collaborative workspace where every relevant person - the vendor, the finance team, the approver, the department head - can see the same thing at the same time and act on it.

Feldman came to this insight with an unusual combination of credentials. He studied economics at Ben-Gurion University in Israel, got his MBA from the Solvay Brussels School of Economics and Management, then spent several years at Colgate-Palmolive implementing SAP ERP systems across global subsidiaries. He understood enterprise software from the inside - the logic of it, and the ways it consistently failed real humans doing real work.

Before Stampli, he spent over a decade at Ness SES, where he built the Documentum document management business in Israel from scratch. He knows what it takes to sell complex software to enterprise buyers, and he knows what it feels like when that software lets them down.

Stampli is now one of the most customer-loved products in the accounts payable software market. It processes more than $100 billion in invoices annually. It serves 1,700+ enterprise customers. It's rated #1 in customer satisfaction on G2 and leads nine separate G2 categories for AP automation in 2025. The company has 280 employees and has raised $145.7M in total funding, including a $61M Series D led by Blackstone in October 2023.

What makes Stampli different is not a single feature - it's a philosophy. Feldman built his own ERP integrations from scratch. All 70+ of them. Most competitors outsource this work or use middleware. Stampli does it natively. The payoff is that when an AP manager at a company running NetSuite calls with a problem, the Stampli support person on the other end already knows NetSuite. Because Feldman hired them specifically for that.

That commitment to domain expertise runs through everything. It shows up in the product, the support team, and in Billy - the company's AI copilot for accounts payable, now in its third year of operation. Billy automates invoice coding, flags exceptions, routes approvals, catches duplicates, and answers questions about specific invoices. It's not a chatbot bolted on to a spreadsheet. It's built on workflow data from thousands of real AP processes.

Feldman is skeptical about AI hype in the way that only someone who's actually shipping AI can afford to be. His line on the topic has been consistent since before it became fashionable: claims are cheap; results are what matter.

Stampli's AI does something measurable. According to Feldman, on average 86% of the finance work across 2,500+ unique data fields is now performed by their AI system. Some customers have used Stampli to avoid hiring dozens of additional staff as their invoice volume scaled. The AI doesn't replace AP teams - it removes the repetitive grind so they can focus on exceptions, relationships, and strategy.

Feldman's 2026 vision for the category is more ambitious still. In March 2026, Stampli launched Deep Finance, a product that converts historical invoice data into real-time executive spend intelligence. The pivot from "process invoices faster" to "understand what you're actually spending and why" is exactly the kind of move a founder makes when they've won the first battle and are already thinking about the next one.

He calls this the "human-in-the-loop" philosophy. AI handles the volume. Humans handle the judgment. The goal isn't to remove people from AP - it's to give them back the time to do the parts of their job that actually require a person.

Feldman makes a point that sounds obvious once you hear it: invoice approval is too late. By the time an invoice arrives, the money is already committed. The approval is a rubber stamp with paperwork. Real financial control requires moving the conversation upstream - to the moment a purchase is being considered, not after it's been made.

That insight is driving Stampli's expansion into procurement. In late 2025, the company announced major momentum in procure-to-pay, building workflows that connect the purchase request to the purchase order to the invoice to the payment - without any of the friction that traditionally made procurement software the most hated tool in the office.

Feldman's argument is that procurement software has historically failed because it was designed for compliance, not for the person making the purchase. Stampli's bet is that if you make procurement conversational and intuitive, the resistance disappears - and the CFO finally gets the spend visibility they've been asking for.