Rippling launches AI Spend Console to rein in runaway AI spend
HR software provider Rippling this week unveiled AI Spend Console, which it calls an anti tokenmaxxing product that helps a company track and contain its AI spending. The tool maps how much individual employees, teams and roles are spending on AI and whether that spend produces genuine productivity or just more AI generated slop; one feature flags, in the company's own words, "which engineers have high AI spend whose peers frequently ask them to redo work in code reviews."
The product grew out of Rippling's own crisis. In March, CFO Adam Swiecicki presented the executive team with a number that shocked them: the company was on track to burn 40% of its R&D headcount budget on AI tokens, with spending growing 80% month over month. If that trend continued, the following year AI token spend would reach 90% of what Rippling paid its high paid R&D staff in compensation. "We were incredulous," said Chief Product Officer Matt MacInnis.
An internal analysis found that roughly 10 to 15% of employees were driving about 60% of total AI spend, and one engineer alone was spending $50,000 a month. Rippling did not want to stop AI usage, just rein it in. It negotiated spending caps with Cursor, OpenAI and Anthropic, and discovered employees were defaulting to the newest, most expensive frontier models for every task regardless of need. MacInnis said inference providers such as Anthropic and OpenAI "have absolutely no incentives to help you control your spend" and do not share usage insight or collaborate with one another.
Rippling built its own AI gateway, bundled into AI Spend Console, that routes prompts to the most cost effective model for the task; enterprises that already use a different gateway can still use the product, but need Rippling's own gateway for the features that govern spending. CEO Parker Conrad said internal benchmarking found SpaceX's Grok the strongest all around performer, while Z.ai's GLM 5.2 was, in his words, "85% cheaper but nearly identical performance," a model also championed by Databricks. The dashboards score attributes such as prompts per day combined with work output, meaning lines of code or pull requests, against spend.
The results: token spend fell from 40% of headcount budget to about 15%, without cutting usage. Rippling hit a peak of 605 billion tokens the month of the CFO's warning; internal usage in July hit 600 billion tokens again, yet July's token spend cost only 37% of what April's had, because prompts were now routed to more efficient models. "That's just because now we're routing to the more effective models," MacInnis said, joking that "we're not letting the sales team do grammar updates using Fable."
Rippling also designated its most effective AI users as "AI captains" tasked with helping the rest of the company, though usage remains concentrated among software engineers; the company is now working on extending similar tracking to customer onboarding teams. AI Spend Console ships included for Rippling's HR subscribers, with additional AI usage based costs, and can also be bought as a stand alone product integrated with another HR system of record.
Key facts
- Rippling's AI Spend Console tracks employee, team and role level AI spend against productivity, flagging engineers whose AI generated code peers frequently have to redo.
- In March, Rippling's CFO warned the company was on track to burn 40% of its R&D headcount budget on AI tokens, growing 80% month over month; one engineer alone was spending $50,000 a month.
- Controls cut token spend to about 15% of headcount budget while usage held near 600 billion tokens a month, and July's token cost was only 37% of April's cost at similar volume.
- Rippling built its own AI gateway, bundled into the product, that routes prompts to cheaper models such as Z.ai's GLM 5.2, which internal benchmarks found 85% cheaper with nearly identical performance to frontier models.
- AI Spend Console ships included for Rippling's HR subscribers with added usage based costs, and can also be purchased as a stand alone product.
Why it matters
Rippling's story is a data point for a wider shift: after a year of unchecked tokenmaxxing, companies that let employees loose on frontier AI models are discovering the bill scales faster than the value. Rippling's own projection, that unmanaged spend would reach 90% of its R&D payroll within a year, is the kind of number that turns AI budgeting into a CFO problem rather than an engineering one. The response, an AI gateway plus per employee productivity dashboards, treats AI tokens the way companies learned to treat cloud compute: something that needs routing, caps and visibility, not just access.
Who it affects
Engineering leaders and CFOs at companies with heavy AI coding tool usage are the direct audience, since the product is built to answer exactly the question Rippling's own executives got blindsided by. It also puts pressure on inference providers and coding tools, Cursor, OpenAI and Anthropic among them, whose token consumption is now being tracked and capped rather than left open ended. HR teams evaluating Rippling as a system of record are a secondary audience, since the tool is sold both bundled with HR subscriptions and as a stand alone integration.
How to use it
AI Spend Console is included for Rippling's existing HR subscribers, though it carries additional AI usage based costs on top of the subscription. It can also be purchased as a stand alone product and integrated with another company's existing HR system of record. Getting the full spend governing feature set requires routing traffic through Rippling's own built in AI gateway; enterprises that already use a different gateway can still use the product, per Matt MacInnis, but without those governance features.
How solid is it
The account rests on Rippling's own blog post and a TechCrunch interview with Chief Product Officer Matt MacInnis, so the headline numbers, the 40% budget share, the 80% month over month growth, the $50,000 a month engineer, the 605 billion and 600 billion token figures, are all self reported by the company selling the product, not independently audited.
Risks and caveats
Because Rippling built AI Spend Console to sell as well as to use internally, its case study doubles as a sales pitch, and the figures it highlights are the ones that make the product look most necessary. The source gives no total dollar figure for Rippling's overall AI spend beyond "millions of dollars," no headcount for the "AI captains" program, and no standalone pricing beyond "additional AI usage based costs." Turning per employee AI spend and output into a scored dashboard also raises an obvious tension: the same tool built to curb wasteful spending doubles as individual performance surveillance, and how that lands with the workforce it measures is not addressed in the source.
“The truth is that the inference providers, like Anthropic and OpenAI, have absolutely no incentives to help you control your spend. They have every incentive for it to be a runaway expense, and that's exactly what they do. They don't provide you with great usage insight, and they don't collaborate with one another.”
— Matt MacInnis, Rippling Chief Product Officer