Content Copyright © 2007 Bloor. All Rights Reserved.
Denodo is different for at least two reasons. First, it has Spanish origins (its technology is based on research started at the Technical University of Madrid) though it is now headquartered in the United States. Secondly, and more significantly, as a vendor of a federated platform for enterprise information integration (EII) it has taken a different path from its rivals. It is the latter difference that I want to discuss.
Denodo was founded in the late 90s with data federation technology at around the same time as Ipedo, Composite Software, and various others that have since been acquired by larger vendors. However, while Ipedo, Composite and the various acquirers have focused on developing and extending their ability to reach out in more efficient ways to internal data sources that they can query (or search in the case of content management), Denodo has also had an additional focus: on providing the ability to federate with third party data as well as with internal data. In particular, with the release of version 4 of the Denodo Platform the company is now targeting what it refers to as “Enterprise Data Mashups”, which is real-time data integration for structured, unstructured and Web sources.
The Denodo product set is actually modular. If you want to do just conventional EII stuff then you can use just that, but for the full power of the product I need to explain what else you can do. Briefly, this starts with web extraction and automation. What this means is that you can automate the process of monitoring, navigating and extracting information from web sources such as blogs or sites where articles such as this are posted, or competitor sites, or whatever. Moreover, the software can search and index against this information, where indexing effectively means applying a structure to what otherwise is unstructured or semi-structured information so that you can query it in a conventional manner. You can also do the same with unstructured content such as email, Word and pdf documents, notes fields in applications and so on. You could also plug in Babelfish or something similar if you want to filter by language. Workflow (which is graphical) is provided in order to help you define the automation process.
These disparate data sources can then be normalised and related using common metadata and semantics. The product supports two sets of users, at least in terms of those making queries. Power users work in VQL (virtual query language), which is an extended version of SQL that supports hierarchical XML (XML is stored this way) and semantic operations like textual similarity matching. End users, on the other hand, will make use of the Query Builder, which effectively provides a 4GL-like environment on top of VQL. Here you can do things like create virtual joins using graphical techniques. The resulting “data mashups” can be exposed as SOA data services.
As you can imagine this ability to link Internet and internal resources, whether they are structured or unstructured, in a combined query and search environment could be very powerful in the right circumstances—such as combining online customer reviews with internal warranty data, or competitor pricing/marketing offers into a CRM/sales automation tool. As far as I am aware, no other vendor has this capability right now. Business Objects’ recent announcement of the acquisition of Inxight suggests that it may be headed in this direction but it will take a while to integrate Inxight, which provides the web/text/federated search half of the equation, with its existing EII capabilities that it acquired from Medience. Nevertheless, this clearly validates Denodo’s position and, no doubt, other suppliers will enter the market. However, for the time being, Denodo has a clear lead on its competitors. Indeed, one could argue that it doesn’t actually have any competitors right now. I expect that it will have in the future but rivals will have their work cut out to catch up with Denodo.