WhereScape has just started twittering. To look just go to http://twitter.com/wherescape, to follow us you will need to sign up with Twitter. We only started yesterday so not too many tweets yet, but we will see how much fun we can have with 140 characters.
WhereScape RED’s “sweet spot” has long been considered the rapid delivery of projects. Our customers are commonly under extreme business pressure with short time lines. While the pressure hasn’t eased - you could certainly make a case that it has got a lot worse for a lot of our clients - we have noticed that we are doing more and more projects within an infrastructure world.
In a project world WhereScape RED IS the infrastructure: we know about the entire data warehouse, information about all the objects is in our meta data, self documentation and impact analysis covers everything from source systems through to cubes, hopefully all the code is generated, changes can be easily propagated etc etc and life is a beautiful thing.
In an infrastructure world WhereScape RED is a part of rather than the entire infrastructure. A common manifestation of this (and one we are seeing a lot of now) is where we are bought in to provide a rapid data marting capability.
The enterprise data warehouse (EDW) team may be functioning as an excellent governance/data management utility, but there are still local constituencies that go off and do their own thing using the EDW as the base. The reasons given are numerous: it takes too long for the EDW team to make changes for me, the warehouse models are too complex, too much (or too little or not at the needed grain) data, I need slightly different data for my project etc etc.
After all the effort that has been put into building the EDW, it seems a pity to let anarchy reign at the project level. WhereScape RED etains the ability to build rapidly while still providing the EDW team with a manageable environment. An example of where this occurred was a telco who recently purchased WhereScape RED. They had built a pretty good EDW on a Teradata, but a set of users needed to build a churn model quickly. Rather than extract the data into a local data mart (which would almost certainly have been SQL Server) they were able to build a churn model specific to the business unit, on the Teradata platform, preserving the architectural integrity and creating a manageable solution quickly.
We also get involved where customers want to start with a project (as this will free the funding) but want to ensure that whatever they build can be delivered quickly, and be retained as part of the eventual EDW.
This approach has the benefit of providing immediate value, while still supporting architectural value. The WhereScape RED “project” layer is retained, while the source to project ETL is replaced with source to EDW ETL.
Obviously this is easiest where WhereScape RED is used for the conventional ETL as well (as all the meta data can be retained) but it is generally trivial to rewrite the logic in a traditional ETL tool.
We expect to see a lot more of WhereScape RED in an infrastructure world. Data warehouse teams are expected to do more with less, and the idea of retaining the existing data warehouse while using smart software to rapidly deliver project based solutions is compelling.
One the Easter long weekend, Wayne, Steve and another WhereScape hanger on (there’s always one or two around) were in Taupo in the middle of New Zealand’s North Island and decided to go trout fishing.
Taupo is New Zealand’s largest freshwater lake which is central to numerous adventure activities and has a back drop of not one, not two but three volcanoes and has some of the best trout fishing in the world
Of course we could have trekked for hours through dense bush and had a go at fly fishing on a river in the middle of nowhere or if you are a touch lazy you can hire an expert with a boat to take you out on Lake Taupo itself.
We (of course) chartered the comfortable Stratus launch and our skipper Marty delivered the trout fishing goods with more than ten trout landed and six of those were over the legal length limit.
The tea, biscuits and boy banter were most excellent and Marty proved to be a bit of a “Gandalf of the lake”, with no place to hide for trout through his years of fishing experience.
All trout were safely smoked and eaten. A great morning in paradise….
We often get asked where WhereScape RED fits into a BI software stack. When we first designed the product we did not constrain ourselves with fitting into existing architectures. Instead we looked at what we would need to provide in order to speed up development and redevelopment of a data warehouse. The downside is that it meant we created a product that cuts across traditional silos - we create tables and indexes (DBA role), generate code (developer), do transformations (ETL guy), document it (no one ones up for this one) …
WhereScape RED is a development environment that builds data warehouse objects in a relational database (SQL Server, IBM InfoSphere, Oracle or Teradata). We use the database’s loaders to load data, and procedural SQL to transform, join, and aggregate data. We store meta data about what is done, and use this meta data to produce documentation on demand. The code we generate can be modified and can be run through our integrated scheduler. WhereScape RED can build different styles of data warehouses (normalized, dimensional, cubes etc), which can be accessed through standard, off the shelf query and reporting tools.
Some of our customers already have traditional ETL tools. These can be used to preprocess data, or to build enterprise data warehouses. In the latter case WhereScape RED is used to speed up the delivery of a data mart layer.
This helps satisfy potentially rogue user communities who are quite happy to use an enterprise data warehouse as the source, but will not put up with any delay in getting their requirements met. Traditionally these users build (hack?) solutions in their tool of choice (excel, access, cubes etc) outside the control of IT. This is not a major for one off projects, but often these solutions become critical parts of an organization’s decisional infrastructure.
WhereScape RED enables these user communities to still rapidly build out solutions, but with the advantage that they are now be built in an environment that can also be maintained and managed.
