He described the way that SnapLogic can be used to transform data into data services. Data services are essential for Mashups because without data you’re basically integrating applications and thats a lot of heavy lifting. With data services, you have the data in an easily consumable form where it can be processed, mashed up and reformatted in whatever way is interesting.
One of the worst kept secrets in the Mashup universe is that in spite of all the activity around mashups, there are surprisingly few sources of real data. Sure there are RSS news feeds, and a scattering of publically available data sources like crime stats, State Parks, etc. but not a whole lot of data that can make mashups really useful.
There are several approached to overcoming this problem that range from web scraping, to shoe-horning everything into an RSS feed. We at SnapLogic think we have a truly elegant approach to this problem that, if widely adopted, could make virtually every data source easily mashable.
Today, Mike showed a mash up of Salesforce.com sales pipeline data with QuickBooks Accounts Recievable data.
It’s a simple mashup that I call the ‘Deadbead Finder’.
First, you select from the drop down menu a sales rep to analyze. For each rep, the mashup will query QuickBooks and list the invoices that are still outstanding for Accounts for the selected rep. You can examine the date of the invoices to see if any are past due (a more complex mashup would highlight them automatically, but we haven’t done that yet). When an Account is selected, a query is sent to Salesforce.com and the Opportunities for that Account are retrieved. The mashup tells you if there are pending deals that are coming from Accounts that have not paid for their prior purchases.
It’s not rocket science, but it is useful.
We start with a SnapLogic server that is configured to access the data from Salesforce.com and QuickBooks. For SalesForce we’ve written a SOAP Component that uses SalesForce’s WSDL and accesses the data through their SOAP interface. For QuickBooks, we’ve written a Component that accesses the data through their COM interface.
This SnapLogic server presents a raw JSON data feed that can be used directly in a mashup as shown below.
The great thing about SnapLogic Data Services is that they are easily mashable by other SnapLogic servers. In this example, we modified the raw data feeds with an additional filter implemented in another SnapLogic server as indicated below.
This downstream filter is easily implemented by constructing another pipeline that uses the raw data service as an input and applies whatever further processing steps are necessary.
The output of this new pipeline can be consumed by a browser in a mashup, or processed again downstream by any SnapLogic server.
This is just one of the many ways SnapLogic can be used in innovative ways to solve real world integration problems. We’ll be describing some more of these in the days and weeks to come.