Workday's finance team used its own AI. A 5-day forecast now takes under one.
Workday published a film and a customer story on 8 October about running its AI on its own staff first, "long before it reaches a customer", as the story puts it. In the film, Julie Gonzalez, who leads financial planning at Workday, says: "Our headcount forecast that used to take almost five days can now be done in less than a day."
The written case study, published the same day, describes an eight-day monthly cycle and says the team "is shrinking" it to a single day, with AI running the first baseline models in minutes.
Workday also reports time savings in recruiting. It says screening and scheduling tools gave recruiters up to four hours a day back, and over 24,000 hours in the first nine months. By its count, 94% of its staff use its assistant, Sana, week after week.
Neither source says whether the one day covers reconciliation and sign-off or only the first model, how accurate the faster forecast is, or how many people work on it. These are Workday's figures about a product it sells.
🎥 Video: Workday, official YouTube channel · narration: AI voice
What the captions say
- Workday ran its own AI on its finance team.
- The monthly headcount forecast took almost 5 days.
- Now it takes less than a day.
- Its written case study says 8 days, not 5.
- Recruiters got up to 4 hours a day back.
- Over 24,000 hours in nine months, Workday says.
- Workday sells the same AI to customers.
- These are Workday's own figures.
