← All organizations

Finance operations and spend management

Ramp

Ramp provides a finance operations platform combining corporate cards, expense management, accounts payable, travel, procurement, and accounting automation. Finance teams can set spending limits, route purchases for approval, match invoices to purchase orders and receipts, and sync transactions with accounting systems. Employees can submit receipts through messaging apps and book travel within company policy. Its Router.com product gives developers a single API for multiple AI models, routing requests by cost and required performance while connecting model selection to AI spending controls.

Founded in 2019 by Eric Glyman, Karim Atiyeh, and Gene Lee, Ramp named Glyman and Atiyeh co-CEOs in 2026, with Applied AI leader Rahul Sengottuvelu becoming CTO. Its engineering work includes Ramp Inspect, an internal coding agent used to maintain Ramp Sheets. Generated monitors trigger investigations that reproduce bugs in sandboxed environments and propose fixes; engineers review code before merging. Router also evaluates models against production engineering tasks through Ramp SWE-Bench, tying model selection to practical workloads.

As of June 2026, the company reported 70,000+ customers, over $1 billion in annualized revenue, positive free cash flow, and $200 billion in annualized purchase volume. Its customers included Shopify, Notion, and Visa. Ramp raised $750 million in Series F financing that June at a $44 billion valuation, bringing total equity financing above $3 billion.

ramp.com

2 talks

Newest first

2 speakers at AIE

Affiliations reflect their AIE appearances, not necessarily current employment.

Start here

  1. Scaffold Wisely

    Start here for a concrete comparison of ways to handle arbitrary financial CSV files, followed by an experimental email client that extends the discussion to software interfaces.

    Rahul SengottuveluAI Engineer Summit 2025

  2. How Forward Deployed Engineering is done at Ramp

    Start here to understand how enterprise requests move toward build decisions and which stages Mehr proposes supporting with agents.

    Leo MehrAI Engineer World's Fair 2026

Messages from the stage

Executable verification with less scaffolding

Sengottuvelu contrasts deterministic integrations and LLM-augmented pipelines with an agent using a code interpreter and executable verification. The comparison grounds his argument for general-purpose models over brittle, specialized scaffolding.

Test the need before implementation

Mehr uses SAP S/4HANA integration and unnecessary Android implementation examples to examine customer context, alternatives, and cross-customer impact. His Slack-to-Notion workflow places request gathering and scoping upstream of specifications and implementation.

Affiliations reflect each recorded session, not necessarily current employment.

Company sources · checked 2026-08-28