The best of statistical forecasting for your business
— In a world in constant and fast changing, the forecaster is increasingly asked to analyze and help understand the business context. The exponential growth of volumes and data sources can help answer these questions while introducing a real technical and organizational challenge.
Artelys Crystal Forecast allows you to generate forecasts or scenarios of extreme reliability on short, medium and long-term horizons. It becomes very easy to automatically use your data feeds, generate forecasts and increase your knowledge of the future!
why Artelys Crystal Forecast ?
Comprehensive
In your information system, Artelys Crystal Forecast becomes the cockpit of the forecaster.
This is where the forecaster organizes all the stages of the forecasting process, from the data processing to the dissemination of the result, and where he finds all the necessary tools for the production of advanced statistical analyzes.
Powerful
Artelys Crystal Forecast was developed by and for its time series prediction experts who wanted to capitalize their experience and gather their statistical modeling skills in a unique and powerful software.
ACF adapts to datasets with aggregated spatio-temporal or typological granularities as well as large, heterogeneous datasets with fine temporal granularity.
Ergonomic
Organization in sequence of tasks to follow the forecast procedure
Connection to your information system
Generation of automatic imports, from your databases or external sources
Simplified export of your results
Provision of a library of tools and models
Possibility to integrate custom scripts
Adjustable
The user interface as well as the workflows are fully configurable so that variables, statistical indicators, visualization screens as well as modeling and parameterization tools can meet your specific business needs..
powerful models
— A whole range of models available
Powerful models, providing forecasts of extreme precision
Models proven in many numerical experiments in business context
— A whole range of models available
Powerful models, providing forecasts of extreme precision
Models proven in many numerical experiments in business context
— Great possibilities of parametrization to adapt to your context
Choose the type of model : linear or non-parametric multivariate regressions (splines), autoregressive models (S)ARIMAX, neural networks, models with unobservable components, etc.
Choose the explanatory variables
Calibrate the parameters associated with the model structure
— Great possibilities of parametrization to adapt to your context
Choose the type of model : linear or non-parametric multivariate regressions (splines), autoregressive models (S)ARIMAX, neural networks, models with unobservable components…
Choose the explanatory variables
Calibrate the parameters associated with the model structure
— Or your own model
Import your own models using scripts
Compatibility with R software: most recent and robust calibration and projection algorithms of the academic community
— Or your own model
Import your own models using scripts.
Compatibility with R software : most recent and robust calibration and projection algorithms of the academic community
— Complete tools to validate your model
Statistical indicators for the model validation : statistical criteria (MAPE, RMSE), residue evaluation, autocorrelation analysis
Synthetic table representing the model quality:
- Overall quality of learning
- Predictive power
Help messages adapted to accompany you in your model updates
— Complete tools to validate your model
Statistical indicators for the model validation : statistical criteria(MAPE, RMSE), residue evaluation, autocorrelation analysis
Synthetic table representing the model quality :
- Overall quality of learning
- Predictive power
Help messages adapted to accompany you in your model updates.
a software that adapts to your data
— Define the basic structure of your data:
Geographical Analysis
Generation of forecasts at any geographical scale, from from the region section to the national and transnational level
Season, calendar and climate effects
Visualization and materialization of their interactions thanks to an automatic indexing
Effects of exogenous variables
Quantification, classification and visualization of interactions between the main effects (cross analysis, identification of variances and covariances, etc.)
Segmentation of the client portfolio
Artelys Crystal Forecast integrates your current operational/business segmentation and offers to complement it with quantitative clustering
Bottom‑up modeling
Representation of scenarios of techno-economic evolution
eg: Typology / Usage / Equipment segmentation to understand energy demand (long-term modeling)
easy and efficient construction of your DATA projects
— Manage your data rigorously
Automatic imports from your own databases or external data sources
Implementation of a relational structure:
- Representation of the relational links between your variables
- Reconciliation at each aggregation node
An available and editable script library to correct, process or transform your data
Centralization of data in Artelys Crystal Database Manager: historical, processed, planned or backcasted data, as well as all your explanatory variables
— Create and organize your task sequences to define your forecasting process
Data management and analysis
Model calibration
Forecasts and scenarios generation
Visualization and reconciliation
Visualization of the flow of the forecast procedure
Automation of repetitive and time-consuming tasks to save time for analysis and modeling
Use of the library scripts or your own scripts for processing and correcting data
Creation of reports (pdf or excel) or web publication of a dashboard summarizing the results
Automatic export to CSV/Excel or other Crystal applications
Visualization of the flow of the forecast procedure
Automation of repetitive and time-consuming tasks to save time for analysis and modeling
Use of the library scripts or your own scripts for processing and correcting data
Creation of reports (pdf or excel) or web publication of a dashboard summarizing the results
Automatic export to CSV / Excel or other Crystal applications
Artelys Crystal Server can be used, in a complementary way, within a client-server architecture to plan and automate the various operational procedures or to deport the heavy computations associated with the constant growth of data volumes.
Artelys Crystal Forecast can also be linked to web interfaces to publish results, dashboards or reports.
countless applications
— Improve your risk management procedures by reducing your short-term errors
Make your buying operations in the marketplace efficiently and cost-effectively
Artelys Crystal Forecast can be use by any market player who buys or sells power and is financially responsible for deviations in the network it creates.
Our solution allows positioning on intraday, day-ahead and forwards contracts by minimizing forecasting errors that can lead to significant deviations from actual results and represent significant operational costs.
Anticipate operational risks and failures
In regulated contexts, the anticipation of tension between demand and energy supply is an essential dimension of day-to-day network management. Artelys Crystal Forecast supports you in anticipation of these potential defects related to the combination of simultaneous effects and participates in the implementation of appropriate procedures such as demand response or temporary reorganization of the network and allows you to avoid triggering expensive procedures.
— Improve your risk management procedures by reducing your short-term errors
Make your buying operations in the marketplace efficiently and cost-effectively
Artelys Crystal Forecast is used by any market player who buys or sells power and is financially responsible for deviations in the network it creates.
Our solution allows positioning on intraday, day-ahead and forwards contracts by minimizing forecasting errors that can lead to significant deviations from actual results and represent significant operational costs.
Anticipate operational risks and failures
In regulated contexts, the anticipation of tension between demand and energy supply is an essential dimension of day-to-day network management. Artelys Crystal Forecast supports you in anticipation of these potential defects related to the combination of simultaneous effects and participates in the implementation of appropriate procedures such as demand response or temporary reorganization of the network and allows you to avoid triggering expensive procedures.
— Represent medium-term uncertainty on the evolution of demand
For the management of hydraulic reservoirs, storage capacities or “take or pay” contracts, a medium consumption scenario is not sufficient to define suitable medium-term strategies.
With Artelys Crystal Forecast, you can simulate hundreds of scenarios from temperature scenarios and economic evolution assumptions to test the robustness of your system and adapt it.
— Create long-term demand scenarios for strategic investments
The projection of energy demand, subject to profound structural changes in terms of efficiency, flexible technologies or mix of uses, all combined with macroeconomic assumptions is a subject of prime importance for strategic planning at the horizons of 10, 20 or 30 years.
Artelys Crystal Forecast allows you to describe and implement a bottom-up structure of your data and to configure evolution scenarios to perform technical-economic simulations and then use them as input data of a capacity optimization or a strategic study on the impact of policy decisions.
— Create long-term demand scenarios for strategic investments
The projection of energy demand, subject to profound structural changes in terms of efficiency, flexible technologies or mix of uses, all combined with macroeconomic assumptions is a subject of prime importance for strategic planning at the horizons of 10, 20 or 30 years.
Artelys Crystal Forecast allows you to describe and implement a bottom-up structure of your data and to configure evolution scenarios to perform technical-economic simulations and then use them as input data of a capacity optimization or a strategic study on the impact of policy decisions.
— Managing the evolution of the behavior of your client portfolio
Any company offering a service to a customer portfolio needs to anticipate its structural evolution and determine the consequences on consumption in order to better adapt resources and means of supply.
Application example:
- By a telecom company to monthly order the number of resources to allocate to call centers taking into account the diversity of requests and their seasonality
- By medium and large-scale retailers wishing to have precise forecasts of sales by type of product or purchasing sector, for each segment of their portfolio in order to improve their reporting and financial planning and optimize inventories
download the brochure
Request demo
