Speed
A full set of forecasting models in minutes.
Gazelle Forecasting Optimizer
Gazelle Forecasting Optimizer is a fully automated, multi-industry forecasting solution. It tests a wide range of internal and external variables, selects the ones that matter, and validates every forecast it produces.
The problem
Many forecasts extend past sales forward. They miss the drivers — price, promotion, seasonality, market conditions — and tuning better models by hand takes analysts weeks.
History-only forecasts miss the reasons demand moves.
Choosing which variables to include is slow, manual and subjective.
Models are rarely re-tested once they are in production.
Forecasts are hard to connect to the business decisions behind them.
How it works
Every stage is automated, so the search for the best model happens in minutes instead of weeks.
Hundreds of Random Forest models run automatically to identify the most important causal variables.
An automated regression-based algorithm measures how each driver moves demand, running hundreds of iterations to find the best-fit model.
Thousands of models forecast the drivers themselves — chosen to maximize fit and minimize out-of-sample error.
The final forecast is calculated from the forecasted drivers, with automated insights on what is behind it.
Hundreds Random Forest models
select the causal drivers
Hundreds regression iterations
measure each driver’s impact
Thousands driver forecasts
max fit, min out-of-sample error
1 validated forecast
calculated from the drivers
Fully automated — minutes, not weeks.
What it evaluates
How well the model explains the demand you have already seen.
Accuracy on data the model was not trained on — the honest test.
Which internal and external variables matter most, ranked.
How much each variable moves demand, measured by regression.
Why it’s different
A full set of forecasting models in minutes.
Drivers are selected and forecast automatically to arrive at the optimal algorithm.
Scales to businesses of any size, across industries.
Your team runs it — no dependency on external vendors.
See it in action
We show the real product on real data in a 30-minute walkthrough — not a mock-up.
Data & integrations
Internal drivers
External drivers
Systems
Demand and revenue across industries. The level of detail and forecast horizon are set during onboarding to match how your business plans.
A wide range of internal and external variables can be tested. The AI selects the most important ones automatically, so you don’t have to decide in advance.
Models are chosen to minimize out-of-sample error — accuracy on data the model has not seen — not only to fit the past.
Yes. Price and promotion plans, including the calendar from the Promotion Optimizer, can feed the forecast as drivers.
Related solutions
Promotion planning and full calendar optimization for a higher return on every dollar of trade spend.
Learn more →Elasticity-based pricing decisions for revenue and profit growth — including shifts in competitor pricing.
Learn more →Ask unstructured business questions in plain language. Interpretation, insights and implications are generated automatically.
Learn more →
A 30-minute walkthrough of Gazelle on a category like yours.