Rippling this week launched AI Spend Console, a product designed to help companies track and control AI spending. The tool maps how much individual employees, teams, and roles spend on AI tokens and attempts to correlate that spending with productivity, such as whether engineers with high AI spend produce work that peers frequently have to redo in code reviews. The product grew out of Rippling‘s own experience earlier this year, when it embraced AI aggressively and quickly saw costs spiral. CFO Adam Swiecicki presented figures in March showing Rippling was on track to burn 40% of its R&D headcount budget on AI tokens — millions of dollars. Spending was growing 80% month-over-month, and extrapolated that would reach 90% of R&D compensation within a year. Analysis revealed that roughly 10–15% of employees drove about 60% of total AI spend, with one engineer spending $50,000 a month.

Rippling‘s response was not to ban AI but to rein it in. It negotiated maximum spending caps with providers like Cursor, OpenAI, and Anthropic. It also discovered that employees defaulted to the most expensive frontier models for all tasks. Chief Product Officer Matt MacInnis noted that inference providers have no incentive to help control spend and do not provide good usage insight. The company found it needed multiple models from different labs at various price points, including cheaper open-weight options. In its own benchmarks, Rippling found Grok was the all-around leader but GLM 5.2 was 85% cheaper with nearly identical performance. This led Rippling to build its own AI gateway that routes prompts to the most cost-effective model, which is part of the AI Spend Console product. The console includes dashboards that score attributes such as prompts per day, work output, and spend.

With these tools, Rippling cut its token spend from 40% of headcount budget to about 15%. Token usage remained high — 605 billion tokens in April and 600 billion in July — but the July cost was only 37% of April’s cost, thanks to routing to more efficient models. Technology alone was not enough. Rippling identified effective AI users and made them “AI captains” to assist others. Extending these practices beyond engineering is a work in progress; customer onboarding teams are being tested with AI automation for mailing data and reconciliation, with productivity measured by customer onboarding volume.

MacInnis emphasized the need to link token consumption in G&A and customer-facing functions back to productivity. If that link cannot be established, broader employee access to AI may be restricted rather than becoming as ubiquitous as Slack or email. AI Spend Console is included for Rippling‘s HR subscribers, with additional AI usage-based costs, and can also be purchased as a stand-alone product integrated with another HR system of record.