Ramp stopped hiring for a quarter earlier this year while its revenue was growing more than 100%, co-founder and CEO Eric Glyman wrote on X (Glyman on X).
In the second quarter, 7,000 pull requests at Ramp came from non-engineers, mostly its customer-experience team. Two years ago that number was zero, he wrote.
Business Pill · Operating Leverage
An evergreen Business Engineer explainer on operating leverage. It teaches the general idea only and does not discuss Ramp.
The key insight: Ramp paused hiring while revenue grew more than 100%, then had non-engineers open 7,000 pull requests in a quarter, per its CEO; Glyman puts the bottleneck in organizations, not models.
What Glyman Said
His account: “AI was breaking Ramp”. A handful of people got “100x more productive” and whole functions fell behind them.
The teams using the least AI asked for the most headcount, he wrote. It got bad enough that Ramp “stopped hiring for a quarter”.

What Was Posted on X
Ramp co-founder and CEO Eric Glyman wrote on X on 18 November that Ramp stopped hiring for a quarter while revenue grew more than 100%, and that in Q2 non-engineers, mostly its CX team, opened 7,000 pull requests. The quotes in this piece come from this post.
Why Staff Could Not Use It
Glyman says the problem was not the models. Staff could see what AI could do but could not let it act.
When Ramp told someone in sales to use Claude Code, the first question was “what’s terminal?” Then: could it see the company’s data, Slack and customers? “no, no, and no”, he wrote.
What Ramp Changed
So Ramp “connected it to everything”. A customer emails about a bug, the fix ships, and the customer gets a reply the same afternoon, per Glyman.
Ramp’s own coding agent was already doing engineering work: we covered Ramp’s Inspect agent in September, and its SWE-Bench v2 benchmark this week.
The Wider Stall
Glyman also wrote that business spend on AI tokens grew more than 5x from December to August, then “gone a bit sideways” even as models got cheaper.
His reading: “model progress didn’t stall. organizations did.” That matches the flat spend we tracked in Ramp’s AI Index in October.
The Company Behind It
Y Combinator describes Ramp as a $60 billion platform with more than $2 billion in annual revenue (YC on X). That puts it at no more than 30 times revenue, by our arithmetic.
The post accompanies Glyman’s interview on YC’s How to Build the Future with Garry Tan. Ramp and Glyman’s earlier startup, Paribus, went through YC’s Summer 2015 batch.
The Structural Read
As we read Glyman’s post, the constraint he describes is access, not intelligence: staff could see what AI could do but could not connect it to data, Slack or customers.
As we read it, the 7,000 figure is a change in who ships code at Ramp, not only how fast engineers work.
Eric Glyman, Ramp co-founder and CEO, on X
“the advantage isn’t access to intelligence anymore. it’s being willing to let it do the work.”
Three Implications
FOR OPERATORS Glyman says connecting AI to company data, Slack and customers is what unlocked non-engineer pull requests at Ramp.
FOR HIRING PLANS Ramp paused hiring for a quarter while revenue grew more than 100%, per its CEO.
FOR AI VENDORS Glyman says business token spend went sideways after August even as models got cheaper.
The Business Engineer Lens
This story maps onto the Business Engineer framework Enterprise AI: From Software to Substrate.
The analysis puts it this way: “The product is the interconnection.”
As we read Glyman’s post, Ramp’s unlock was connecting AI to its data, Slack and customers. The analysis predates his post and does not discuss Ramp.
What Is Not Established
Glyman did not say how many people Ramp employs, how long the freeze ran beyond a quarter, or whether hiring has resumed. The 7,000 figure counts pull requests, not shipped features.
The token-spend trend is his statement; he did not name its data source in the post.
The Bottom Line
Ramp paused hiring for a quarter while growing revenue more than 100%, then had non-engineers open 7,000 pull requests in Q2, up from zero two years ago, per its CEO.
97,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. The hiring pause, the 7,000 pull requests, the token-spend trend and the quotes are from Glyman’s post on X. The $60 billion and $2 billion figures are from Y Combinator’s post. The valuation-to-revenue multiple is our arithmetic.
Sources: Eric Glyman (Ramp co-founder and CEO) on X, 18 Nov 2026 · Y Combinator on X, How to Build the Future episode, 18 Nov 2026









