Autobox Cloud Collaborative Forecasting

4/2/2014 Autobox Cloud Collaborative Forecasting Launch!


It’s our new Web based Collaborative Forecasting Solution.  It connects Autobox to and from databases.  It connects people to Autobox in the S&OP consensus forecasting process.  It connects views of historical MAPE accuracies over time.


  • Cloud Based
  • Collaborative
  • MAPE Tracking
  • Parallel Processing
  • Integrated with Databases


Autobox is simply the easiest way to forecast. Built around what has been identified in the “Principles of Forecasting” Textbook(p. 671) as the best dedicated  forecasting engine, it provides a start to finish environment designed to make forecasting easy whether you have one series or one thousand. You can specify your own model or run in a batch environment, but we like the automatic results.  You can tweak the automatic system easily and check your accuracy using our professional forecasting diagnostics by evaluating values to create rolling forecasts. No matter what method you currently use to forecast, Autobox will improve your ability to forecast accurately.

How Cloud Collaborative Forecasting can be used


Forecast incorporating future expected events(ie price, promotion) directly into the model and NOT an adjustment after the baseline forecast:

Daily Demand Plan

Short-Term and Long-Term Demand

Financial Planning


Data Cleansing

Correct historical data to what it should have been due to misreporting or removing the impact of unexpected events (e.g. outliers).



Evaluate historical data to determine if a causal variable(ie promotion) is important and what is the exact time delay or lead impact.


What makes Autobox so Unique?

  • An expert forecasting system that tries to detect patterns within the data instead of "fitting" a model.
  • Other forecasting tools take a set of models and run a "pick best" tournament to fit the model instead of building a customised model for each problem.
  • Incorporates automatic outlier detection
  • Automatically senses relationships like stocking up before a holiday or a drop-off in demand after the holiday.
  • Automatic data-driven adjustments for fixed days of the month e.g. Friday before and Monday after a holiday that falls on the weekend or specific days of the month.
  • Ability to detect and adapt to major changes in the data:
  1. Changes in Level
  2. Changes in Trend
  3. Changes in Parameters
  4. Changes in Variance
  5. Changes in Seasonality
  • Allows the user to obtain and use forecasts for user-suggested predictor variables.

Automatic Forecasting Systems, Inc.

A world renowned company that has revolutionized how time series data is analyzed, modelled and forecast since 1976.

Go to top