Datawatch and data preparation

Datawatch has released its Managed Analytics Platform (MAP). It is somewhat difficult to categorise this as a product. Ostensibly it is an integrated self-service data preparation and business intelligence platform. However, it is somewhat different from other products in this category. In particular, MAP primarily represents the integration of Monarch (the core engine provided by Datawatch) and Panopticon, the data visualisation engine acquired by Datawatch in 2013. You have to know a bit about these products to understand where MAP is different.

To begin, with consider Monarch. Historically, this was a business intelligence product that extracted information from documents and reports that were generated by ERP and other applications. This data could be collected, aggregated and analysed using Monarch. Now, Datawatch claims that what it has been doing all along with Monarch is “data preparation for documents”. They have a point: it’s not quite data preparation in the sense that you might normally think of it, but it is certainly a potential requirement and one which other vendors in this space cannot usually meet. On the other hand, Monarch can also work with conventional data sources such as databases and flat files and can extract data from web pages, so it has a string to its bow that other suppliers don’t.

In addition to this, Monarch added streaming support in its last release and one of the target markets for Panopticon was in visualisation for streaming analytics. So MAP has support for stream processing which, again, is not something you see in other data preparation platforms. The company has also added time series functions to support deployments within the Internet of Things (IoT) and, on a different tack, it has implemented data masking, again something that is lacking from most other data preparation platforms.

So, is this the best thing on the market? Well, not quite. Some 80 built-in functions, derived from Monarch, have built into MAP for data cleansing and transformations, so it should not fall down on that front. However, where it does currently lack functionality is in collaboration. The company plans to introduce facilities to support this, as well as machine learning capabilities, in a later release.

This is sort of the problem in the self-service data preparation market at present. Whether they are stand-alone solutions or integrated with business intelligence engines, all the products on the market lack some capability. Of course, this is typical of an emerging market and, over time, we can expect all the vendors to start to introduce comparable features to those of their competitors. I think Datawatch has an advantage here: it is more difficult to introduce the ability to extract data from a pdf document, or to implement support for streaming analytics than it is to introduce facilities to support collaboration or other of what we might call “nuts and bolts” features.

I have to say that, more generally, I am quite impressed with Datawatch. It used to be something of a well-kept secret but since the company appointed a new CEO (Michael Morrison) and acquired Panopticon the company has become much more visible. MAP will no doubt help that process even further.