Technical Documentation

DEMAND

Demand forecasting with multiple analytical methods

An application for bulk processing of sales data and product-level demand forecasting, with parallel use of different methods and data exchange through web services, Excel or CSV.

  • Demand Forecasting
  • Multiple Methods
  • Excel / CSV
  • Web Services
Indicative DEMAND application view.
On this page

Overview

DEMAND was designed to analyze historical sales data and produce short-, medium- and long-term demand estimates for individual products.

The documentation’s core approach is to allow different methods to run in parallel and compare their results rather than relying exclusively on a single model.

Core capabilities

Bulk processing

Analyze sales data across multiple products.

Multiple methods

Select one or more analytical approaches.

Short-term forecasting

Estimate near-term demand for operational decisions.

Medium-term forecasting

Analyze trends over a broader planning horizon.

Long-term forecasting

Support more strategic demand scenarios.

Parallel comparison

Cross-check results produced by different methods.

Excel / CSV files

Import and export data in common file formats.

Web services

Integrate with accounting or other business systems.

Methodologies

The original documentation records ten selected methodologies that can be used individually or in combination.

  • ARIMA
  • Neural Networks
  • Croston Methods
  • SSA Decomposition Models
  • Adaptive Smoothing
  • Exponential Methods
  • Regression Analysis
  • Recurrence Plots
  • Maximum Entropy
  • Holt-Winters seasonal method

Integrations

DEMAND can receive data from an existing accounting or business system through web services, while the documentation also supports independent operation using Excel or CSV files.

Discuss a related project

Contact AspectSoft to discuss requirements, integrations or the technical approach for a related implementation.