a practical solution for manipulating data before injecting it into Excel

Predicsis & excel.png

 In this blog, you'll discover how AI can help Excel users to save valuable hours finding salient information by automatically proposing relevant correlations, discretization and groupings.



Data analysts are struggling to reconcile the opportunity that new and varied sources of data offers them, with the challenge of analysing that data in a timely, effective manner

The analysis of business data has become central to most organisations’ operations, while the people who use and analyse that data cover a spectrum of capabilities: from the occasional, low-level, “headline” user, to the expert who churns out charts and statistics and really delves into the fabric of the company’s data sources.

With Excel, the most commonly used BI tool, users are able to analyse data in a way that drives faster and more efficient mining of the data available. And, with the variety of data sources now available to companies, the possibilities for deeper, more meaningful analyses are seemingly endless. But with this variety also inevitably comes complexity - a multiplicity of data variables that requires increased time and effort - and skill - to analyse.  Added to this the fact that data sources are often imperfect - missing values, different files and formats - means that finding meaningful insights and correlations can represent a particularly large headache even for some very skilled personnel.

While data analysts are very knowledgeable about their business, it can be frustrating for them to spend great swathes of time “preparing the data” to deliver meaningful analyses.


Greater variety and/or larger amounts of data can be good news if is used to help acts as a data analyst assistant by preparing the data into a manageable format that frees up the analyst to concentrate on extracting meaningful results in a timely fashion.

Feedback we often receive from our users is that one of the biggest challenges when faced with finding meaningful insights from customer data is how to reduce the data in a way that it can then be analysed. The analyst may well already have a reasonable idea of how the data might correlate, yet preparing the groupings for the values and testing out different correlations between multiple variables becomes an all-consuming task, detracting from the more important goal of extracting meaningful, business-directing results.

This is precisely where the PredicSis Automatic Machine Learning tool,, can help users with Excel: simplifies and clarifies the data in minutes, saving the user valuable time, freeing him/her up to analyse data that has already been aggregated and correlated. New data sources can easily and quickly be added to test whether these have a significant impact on the original data and can lead to more meaningful direction for the business, through profiling and Predictive Analytics.


Too much information can lead to overload and business inertia

But it doesn’t have to.  Let's demonstrate the benefits of combining with Excel. Our example data set contains the results of a national demographic survey conducted in 2008, which provides information on 48,000 US citizens, including their individual incomes. The objective is to quickly identify patterns among the different characteristics of the population earning more than $50k/year.

Even a data set of 17 columns with tens of possible values makes the job difficult and time-consuming. In the following video, I show how the automatic data enrichment provided by drastically shortens the process:

We love sharing our work with you

To download your copy of the document, please complete the form and receive your download link.