Developer Preview: xTuple Business Intelligence

 

jgunderson's picture

Progress continues on our next generation Mobile Web client enabling xTuple users to use browsers on both the desktop and mobile devices. The new framework of JavaScript technologies was introduced in VP Software Development John Roglestad's post JavaScript, JavaScript, JS all the way.

What you may not know is that we are also integrating open source business intelligence (BI) technologies in the stack, based on the Pentaho family of technologies. While technologists can argue all day about the distinct nature of transactional systems like ERP and free-standing BI systems, we believe that's a distinction regular users don't really care about. They just want to see their company data presented in an easy-to-read, meaningful and actionable way

Pentaho Business Analytics provides a complete set of open source technologies to develop BI solutions. The suite includes reporting, data access, integration, visualization and data exploration. The suite is also customizable and embeddable allowing us to integrate with the leading edge technologies being used in the Mobile Web client.

Standard Reports

The first step was to provide standard ERP reports through Pentaho Reporting. Think of this part as the Mobile Web version of our OpenRPT report writer and designer.

The Pentaho Report Designer is a visual design tool which makes it easy for authors to create rich reports. The Pentaho Report Engine runs within the Pentaho BI Server and renders and delivers reports in a variety of formats. To integrate with the Mobile Web client, the Report Server uses scripts to request data from the Node.js data source. As you may recall, Node.js is used as the basis for our Web service that both serves our application and handles requests for data.

Analytics

The next step is to integrate business analytics though Online Analytic Processing (OLAP), which allows you to interactively query and view data from different points of view. OLAP also provides key performance indicators (KPIs) to help you measure the particular business metrics important to you. For example, you might compare an item's change in profitability within the same periods in the previous year. And you may want to further segment this by customer and account representative. OLAP query processing is provided by the OLAP Engine which also runs under the Pentaho BI Server.

To allow fast, interactive queries the data used by OLAP is extracted and reorganized. The data is organized to allow historical analysis over large amounts of data. Pentaho Data Integration is used to extract xTuple data, re-organize (transform) it for analysis and load a data mart, i.e., extract-transform-load process (ETL). A more detailed diagram of the BI Server and Pentaho Data Integration is shown below.  Note that the data mart also lives in a PostgreSQL database.

OLAP enables us to use ad-hoc tools for exploring data. Tools allow you to compose tables and charts using drag and drop operations.

Dashboards

Dashboards allow you to quickly and visually monitor the performance of your business. xTuple will supply dashboards that provide an overview using KPIs, comparisons of dimensions over time and operations performance, developed using JavaScript, so you can customize the content and construct your own dashboards. Stay tuned for part two of this blog post, to see how this all looks to the end-user!