FAQ
Common questions about data, models and access
What data do I need to get started?
A sales history with a date, an item identifier and a quantity. Extra columns such as channel, region, price or promotion improve accuracy but are optional.
Which models does the platform use?
A combination of statistical methods, gradient boosting and deep learning architectures for time series. Models are evaluated per series and the best performing combination is selected automatically.
How far ahead can it forecast?
Horizon and granularity are configurable. Daily, weekly and monthly forecasts are supported, from a few periods ahead to a full planning year.
How do I know the forecast is reliable?
Every forecast includes backtests and error metrics compared against a naive baseline, so you can see where the model performs well and where it does not.
Do I need a data team?
No. Business users work through a guided interface. Technical users can inspect models, parameters and metrics whenever they need to go deeper.
How is my data handled?
Data is isolated per organization, encrypted in transit and at rest, and never used to train models for other customers.