Workday to Acquire Honeydew to Give AI Agents One Set of Business Definitions

Workday to Acquire Honeydew to Give AI Agents One Set of Business Definitions

Workday plans to acquire Honeydew, a startup that builds a semantic layer for AI and business intelligence, so that every AI agent and dashboard reading company data works from the same business definitions. The deal was announced on November 18 at the Workday Rising EMEA event in Barcelona, and financial terms were not disclosed. If you have ever seen two reports give two different answers to the same simple question, this Workday Honeydew acquisition is aimed squarely at that problem.

What the Workday Honeydew Deal Covers

Honeydew sells a semantic layer, which is a shared rulebook that says what each business number actually means. Workday will plug Honeydew’s “semantic compiler” into Workday Data Cloud, its platform for HR, finance and IT data. The compiler turns those business definitions into exact, governed database queries.

Workday says the deal should close in the fourth quarter of its fiscal year 2027, which ends January 31, 2027. Closing still depends on the usual conditions. Until then, Honeydew remains a separate company.

DetailWhat we know
BuyerWorkday (NASDAQ: WDAY)
TargetHoneydew, a semantic layer for AI and BI
AnnouncedNovember 18, 2026, at Workday Rising EMEA
PriceNot disclosed
Expected closeWorkday’s fiscal Q4 2027 (ends January 31, 2027)
Where it landsWorkday Data Cloud

Why AI Agents Keep Giving Different Answers

Workday uses a simple example to explain the issue. Ask a dashboard, a chatbot and an AI agent for your company’s headcount, and you might get three numbers. One counts only full-time staff, one adds contractors, and one treats approved but unfilled roles as already staffed.

None of those answers is technically wrong. They just rely on different definitions, and that inconsistency becomes a real risk once agents start acting on the numbers. Honeydew maps several definitions and picks the right one based on who is asking, the use case and the context.

The “schema-stuffing” workaround

Today, many data teams connect AI to company data the hard way. Workday says they write custom code to catch model requests, hardcode business rules into data pipelines and pack huge data dictionaries into every prompt. When a team switches models or tools, much of that work has to be rebuilt.

How the Honeydew Acquisition Changes Workday AI Queries

The core idea is to move business logic out of the prompt. Instead of asking the model to work out how to calculate payroll or revenue, Workday will build the exact query first and hand the model a clear path to the answer. Workday CTO Gabe Monroy put it plainly: an AI model “shouldn’t be improvising the calculation behind a payroll number.”

Workday lists several benefits it expects from the setup:

  • More consistent answers across dashboards, chat tools and agents
  • Fewer tokens per request, since prompts no longer carry full data dictionaries
  • Lower costs and better efficiency, according to the company
  • Definitions that stay intact when customers switch AI models or BI tools
  • Queries that run on live data in Workday Data Cloud instead of older warehouse copies

Permissions travel with the definitions

Governance is a big part of the pitch. Workday already controls agent identity, access and behavior, and Honeydew adds a check at the moment a question becomes a query. Each person or agent sees only what they are allowed to see, and unauthorized requests are blocked before any query runs.

This fits a wider push by enterprise software makers to wrap tighter rules around AI agents. Recent moves include agent governance tools from ServiceNow and the Copilotforce CRM tie-up between Salesforce and Microsoft. Data platforms are chasing the same goal, as seen with Databricks Genie.

What This Means for You

If your company runs on Workday, this deal hints at where its AI features are heading. Agents that answer HR and finance questions should, in theory, start giving the same number your finance team already trusts. That matters most for payroll, headcount and budget figures, where small differences cause big headaches.

Here are a few practical steps to take now:

  1. Write down how your company defines key metrics like headcount, revenue and attrition, since a semantic layer is only as good as those definitions.
  2. Ask your Workday account team when Honeydew features will reach Workday Data Cloud, because nothing changes until the deal closes.
  3. Review which data analytics tools your teams use, and note where their numbers disagree today.
  4. Check how your current AI agent setups handle permissions, so you know what governance gaps exist.

Keep in mind that these are plans, not shipped features. Workday notes that unreleased functions may change or may not arrive at all, so base buying decisions on what is available today. You can follow more enterprise AI moves on our AI news page.

Sources

Primary source: Workday’s Honeydew acquisition announcement. Additional details from the PR Newswire release.

Conclusion

Workday’s plan to buy Honeydew targets one of the quieter problems in enterprise AI: agents and dashboards that disagree because they define the same number differently. By compiling business rules into exact queries before a model ever sees the request, Workday hopes to make agent answers more consistent, cheaper and better governed. The price is undisclosed and the deal should close by January 31, 2027. For Workday customers, the real test will be whether those consistent, permission-aware answers actually show up in everyday HR and finance work.

Newsletter

New AI tools in your inbox

The best new tools and the AI news that matters, in one short email. Unsubscribe any time.

Free · No spam · Unsubscribe any time

✓ Looks like you’re subscribed

Loading Google’s sign-up form…

Looks like you’re already subscribed

Thanks for reading! New AI tools and the AI news that matters will land in your inbox.

Not subscribed yet? Refresh the page to load the sign-up form.

The sign-up form couldn’t load. If you use an ad or content blocker, allow news.google.com and reload the page.

Free instant check before you pay. Takes a few seconds.

Scroll to Top