Connecting to content

I talked about EntropySoft in a couple of articles I wrote in the spring of last year. However, now I want to go into a little more detail, partly because the company has seen significant growth in the number of its OEM partnerships and partly because the company, which was selling connectors, Content ETL and Content Federation to both end users and OEMs is now just concentrating on the latter, and focusing exclusively (unless you ask very nicely) on OEMs and partners.

There are a number of things that set EntropySoft apart. The first is the sheer number of the connectors it offers: approaching 30 at present, including Microsoft SharePoint, IBM Lotus Quickplace, Hummingbird DM, Alfresco, FileNet, Interwoven, Documentum and so on. The next on the list will be an updated version of their EMC Documentum e-room connector, which will be available later this month.

Note that these connectors, which are bi-directional and support both read and write capabilities, have been specially designed for content-centric applications. In particular, they understand the logic that content management systems impose on the content that they are managing and support domain-specific objects such as documents, versions, permissions and metadata, whereas conventional ETL vendors have connectors for structured data or, where they support unstructured data, can only move Word documents (say) from one place to another, without understanding their context.

So far, so good. But if I left this article at this point you still would not understand a major benefit of using EntropySoft’s connectors, which is that they have a high number of features specifically adapted to the different markets where EntropySoft is selling its technology. This is because different markets often require different technology. For instance, if you are a search vendor you will have different requirements from a company in the business process management space, which will be different again from the requirements of, say, the records management market.

As an example, within the enterprise search market the emphasis is very much on the speed of retrieval and the need to be able to rapidly update indexes. Therefore technologies that might impede this process are to be avoided at all costs. Specific features that EntropySoft includes for the search market include repository partitioning based on metadata, features to support index-update time-reduction, the ability to read document access control lists, group membership retrieval, advanced user directory features (search, browse and group membership), support for all object types, and the ability to extract specific objects (news, forums, messages, calendars, tasks, list and so on) for SharePoint, e-room and Livelink.

I won’t go into a comparable list of features for EntropySoft’s other target markets such as business process management but suffice it to say the proof is in the pudding. Search has been the company’s first focus and it has 8 suppliers in this space using its connectors of which the most well-known is probably Endeca; at the same time it has a couple of partners in the business process management space, headed by IDS Scheer; and it has recently signed eDiscovery vendor Kazeon as an OEM. So the company’s strategy is clearly working.

While it continues to expand the number of content stores it supports the company is also working to increase its interoperability to support additional programming environments. Currently, its connectors are already available in Java or via web services and .NET. However, it is working on a new API for C++ environments. In the longer term, the company’s objective is to become the most widely used technology for complex, fully-featured repository-based content access and to become the de facto standard for all document-centric markets. While that may seem a tall order the company has a healthy lead in its chosen market and, as of today, if anyone is going to achieve such a position then it will be EntropySoft.