ETPlogo4

Copyright © 2010 Eighty Twenty Plus Ltd. All rights reserved.

Telephone (USA) +1-937-369-8020

clouds 160x41_Get_Flash_Player

Predictive modelling is a process that results in the discovery of patterns and relationships in data that may be used to make valid predictions.

 

As an example, a model that is built to understand when and why a customer is likely to leave the business (churn), can be used to predict the likelihood of a customer to leave based on changes in their current behaviour.

 

Building predictive models is part of a larger process that includes everything from defining the basic problem the model will solve, to deploying it into a working environment.

What are Predictive Models?

Xeligence Predictive Modelling

Are you tired of paying over the odds for your predictive analytics solution? Having trouble operationalising your predictive models to provide value through your organisation? Sick of waiting for weeks and even months before your predictive models can be deployed?

 

Xeligence is a family of business focussed analytical applications that continuously improve the performance of core business processes. Through our extensive experience of developing predictive modelling solutions we have created Xeligence to enable the business user to easily create and rapidly deploy predictive models in real time, within hours, without the specialist need of a statistician.

 

Most large organisations have spent millions investing in CRM systems and analytics solutions, however the gap often still exists in utilising the data to make decisions and operationalizing the analytical insight. Xeligence automates the model building and maintenance process and the deployment of the resulting models across all channels within the organisation.

Xeligence Flow

The Xeligence Solution

The Xeligence solution automates the process of developing and maintaining end-to-end data mining solutions. This incorporates all ETL from source systems, variable selection and classification, modelling and scoring phases, and deployment and ongoing tuning and maintenance. This enables your company to save thousands of statistician work hours and decrease time-to-market using predictive models for business decisions. Xeligence also incorporates self learning predictive analytics—the predictive modelsels automatically learn from each new data variable or data update by autonomously updating and deploying responses in real time.

Traditional data mining tools typically involve high start up and resource costs, lengthy deployments, require very skilled statisticians to manually build and refine predictive models, have a lack of flexibility to the changing environment, and a lack of real time decision making capabilities. While Xeligence has its roots in data mining, it breaks down these traditional barriers and allows the business to focus on creating value for its customers.

 

As part of the implementation we deploy an industry specific data model (currently Telecom, Financial Services, Retail, Healthcare) and integrate the data into the solution. The analytical engine automatically builds the predictive models based on attributes in the data model, as new attributes and data are added to the data model, the predictive models automatically adjust and self train to create the most accurate business rules to be deployed into every business process. Numerous models can be built very quickly enabling deployment either to different segments, business units, channels, or the geographic locations your business determines.

More ...

Xeligence Architecture

Xeligence is a business solution not a tool – it is driven by the business user not by specialists

The solution is “Delivered integrated” not just “Can be” integrated

Delivered at a fraction of the price of other tools (that don’t include the implementation and   deployment costs)

Has self learning and self training models – optimises, self manages, and automatically recalibrates models on arrival of new data and attributes

Real time extensions allow predictive decision making to be integrated into operational processes

Integrates seamlessly with Teradata (and other vendor) industry LDM’s

Typically delivered in 3 months – generating results fast. Additional application areas even faster

Compresses model building and deployment time from weeks to hours or minutes

Decades of industry experience already built into vertical solutions

Deploy predictive models in real time into operational systems and processes

The recommendation engine becomes the central control of interaction policy for the marketer

The only solution in the market that does not require the use of a statistician – Xeligence is designed for the business user

Xeligence Architecture