ML-Powered Demand Planning
Turn sales data into
forecasts you can act on.
Upload 18–24 months of history. Get demand forecasts, uncertainty ranges, and planning KPIs for inventory & production — in minutes.
WAPE
8.5%
Bias
−2.1%
Accuracy
+91%
New forecast ready!
8 SKUs · Q2 2025
Required files
Sales history
18–24 months of monthly or daily sales by SKU & channel
Product master
One row per SKU with product attributes
Business calendar
Holidays & special events for seasonal pattern detection
Optional files
Promotions & pricing
Promo periods, discounts, and price changes for scenario planning
Inventory signals
Stock levels and out-of-stock flags to improve planning accuracy
What you get
Planning outputs
built for operations.
- Demand forecast 3–6 months ahead with P10–P90 uncertainty range
- Planning KPIs: WAPE, Bias, Under/Over-plan rate, Top SKUs to fix
- Scenario planning for promotions, pricing, and distribution changes
Forecast Horizon
6 months
SKU Coverage
98.4%
Reorder Alerts
Auto
Time to Insight
< 2 min
Forecast models run natively on Dataiku DSS — fully integrated for automated retraining, versioning, and one-click deployment.
How it works
Upload your data
Connect sales history, product master, calendar, and optional files.
Models train & validate
LightGBM + Prophet models run with automatic quality checks.
Explore & export
Review forecasts, KPIs, and scenarios. Export for production teams.