The AI compensation analyst can now work from merit cycle history, a company’s own pay ranges and survey methodology, real-time job posting data, and uploaded documents — surfacing recommendations compensation teams can trace, question, and defend
SAN FRANCISCO, Aug. 13, 2026 /PRNewswire/ — Pave, the AI compensation platform, today announced a significant expansion of the Pave Agent, its AI compensation analyst. The Agent can now reason across a company’s full compensation context: every merit cycle in Compensation Planning, the pay ranges and survey methodology built in Market Pricing, a new real-time job posting dataset, and the customer’s own documents.
Most compensation questions cannot be answered from a single source. Adjusting a range means weighing market movement, survey blend, and recent cycle history at once. Understanding merit budget outcomes requires both cycle and employee records. Judging whether an offer is competitive depends on a company’s bands, its pay philosophy, and what the market is paying today. Until now, connecting those threads meant hours of manual work across spreadsheets and systems.
With this release, that analysis happens in minutes — and the Agent shows its work. Every recommendation cites its sources and explains its reasoning. The Pave Agent remains advisory: it recommends, and the compensation team decides.
“Compensation leaders don’t need an AI that answers questions in isolation. They need one that understands how their organization actually pays people — their philosophy, their leveling, their history,” said Matt Schulman, CEO and founder of Pave. “That’s the difference between a generic AI tool and an agent that was purpose-built for compensation. The Pave Agent brings the analysis. Your team brings the judgment.”
Merit cycles, in full context
The Pave Agent can now reference every merit cycle run in Pave Compensation Planning, as well as employee compensation history. Compensation teams are using it to complete cycle retrospectives in minutes — showing how increases were distributed across departments and where manager recommendations deviated from guidelines — and to bring an employee’s full history across cycles into a single pay decision. Cross-cycle comparison allows assessment of whether budget deployment actually reflected the company’s stated philosophy.
Compensation Planning remains where guidelines are set and manager decisions are made. The Agent adds the analytical layer on top, surfacing patterns, risks, and context for human review.
A company’s ranges, built on its own methodology
The Agent now works from Market Pricing — a company’s survey uploads, data rules, blending logic, and every version of its band sets. Because Market Pricing serves as the compensation system of record, each range carries the methodology behind it: which sources were used, how they were weighted, and what was adjusted.
That means questions like “why is this range what it is?” or “how does our draft Engineering ladder compare to what’s live today?” are answered using the company’s own methodology and decision history—not a generic benchmark.
Asking the market directly
The Pave Agent can now analyze Pave’s public job posting dataset—job titles, descriptions, and posted salary ranges from publicly available listings—enriched with location and job family details and refreshed daily.
Pay transparency legislation has made posted ranges one of the most dynamic signals available, offering a real-time read on how the market describes, levels, and prices roles. Compensation teams are using it to price roles in unfamiliar industries where they have limited history, to sharpen job descriptions to align with how the market defines a role, and to track how posted ranges for a role shift over time.
Posting data is informative, not prescriptive. The Agent presents it as one input alongside a company’s survey blend and Pave’s real-time benchmarks, with sources attached so teams can weigh it themselves.
Bringing your own context
Compensation decisions are shaped by documents that live outside any single platform: pay philosophy, equity plan design, third-party survey files, and offer letters. The Agent now works with those to identify the best response. Files can be attached to a single conversation for immediate analysis or uploaded as persistent context—up to 100MB—so they inform every future analysis, much as external surveys are blended into Market Pricing.
Because compensation data is among the most sensitive information a company holds, file handling was built accordingly. Uploads are secured to the customer’s tenant and separated from other customers. Session files are processed in an isolated environment that is destroyed when the conversation ends. Uploaded data is used only to answer that customer’s questions and is never used to train models.
Availability
The new capabilities are available today to Pave customers.
To learn more about this release: https://www.pave.com/blog-posts/compensation-context-pave-agent
About Pave
Pave is the AI compensation platform, purpose-built for compensation leaders. Through real-time connections to HRIS, ATS, and equity systems, Pave gives teams a continuously current data foundation for benchmarking, building pay ranges, running merit cycles, and communicating compensation to employees. The Pave Agent, Pave’s AI compensation analyst, reasons across that foundation to deliver explainable, advisory recommendations that compensation teams can defend to their CEO, board, and employees. Pave is headquartered in San Francisco. Learn more at pave.com.
View original content to download multimedia:https://www.prnewswire.com/news-releases/pave-expands-the-pave-agent-to-reason-across-a-companys-entire-compensation-system-of-record-302850545.html
SOURCE Pave
