Some interesting news from SAS today that, to me at least, boosts their support of enterprise decision management or EDM. First they announced a specific product bundle from their work with Teradata – you can get details here. This integrates SAS Enterprise Miner, their predictive analytics/data mining workbench, with Teradata’s database allowing for models developed in Enterprise Miner to be deployed to and executed in the Teradata server. Executing the models this way results in a 45x throughput improvement. Yes, 4,500%. Of course the slow execution performance of models on SAS was one of the reasons so many customers used base SAS and Enterprise Miner to develop models but deployed them differently so the huge jump may come partly from a low baseline. Nevertheless this is great news as it makes it easier for SAS customers to push their models into a high-performance execution environment. As EDM requires both rapid deployment of models into operational systems and the execution of those models so that they can be used in a high-volume, quick response system this is good news. Darryl McDonald, Chief Marketing Officer at Teradata said that the partnership would make customers able to make “smarter decisions with speed and precision, giving them competitive advantages in their market”. Precision and speed are two of the measures of a decision but I also see this relationship improving consistency (by running the models on the shared server the same model is used everywhere), boosting agility (by automating deployment models are put into production faster) and lowering cost. Excellent news and I hope to get a more detailed briefing soon.
The other interesting SAS news was the acquisition of Teragram (details here). This is interesting as it reflects the increasing importance of text analytics and search in the analytic space, something that a focus on EDM and thus on operational, front-line systems puts front and center. After all much of the text being analyzed is about customers and customer interactions and so important to analyze if you are really trying to improve customer decisions.
Off to the DAMA conference…