AustraliaFood and restaurants2020 – ongoingDomino's Pizza Enterprises
Cognitive Rostering: machine learning sales forecasts for store rosters
Domino's built store rosters from machine learning sales forecasts in Australia and New Zealand
Problem
Store managers and franchisees set staff rosters largely by guesswork. Stores risked being understaffed on busy nights and overstaffed on quiet ones.
Approach
Domino's worked with the Australian AI firm Max Kelsen to build Cognitive Rostering on Amazon Web Services, using SageMaker, EKS and S3. The model forecasts store sales from public holidays, sporting events, marketing activity, historical data and seasonal changes. Stores use the forecast to build rosters.
Results
No published figures
Takeaway
Domino's tied staffing to a demand forecast that includes events and marketing plans, not only past averages. The company gave no accuracy figure.
Categories
- Size
- Large enterprise
- Industry
- Food and restaurants
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Partner co-development
- Evidence
- Company's own claim
Sources
Source summaries
Domino's announced the tool, built with Max Kelsen on AWS, for stores across Australia and New Zealand. It says the forecasts significantly improve sales prediction compared with historical averages. The release gives no accuracy figure.
Full details are in the original sources.
