I got a briefing from BEZ recently, after Claudia Imhoff (@claudia_imhoff) referred them to me. BEZ has historically worked on the IT side of organizations, using predictive analytics to take historical data about system performance and transactions and use it to predict application response times, throughput, utilization of resources etc. BEZ compares itself to a traffic-enabled GPS for IT departments. It runs 7×24, collects and aggregates data continuously from lots of systems, generates advice and manages alerts about service levels you might miss in the future. It is designed to be easy o install, hands off and automated.
BEZ supports strategic (capacity planning), tactical (activity profiling and performance management through changing indices, SQL statements, JVMs) and operational (workload management and optimization with a business application focus) decisions. A key element is the aggregation of performance data into application-centric workloads. This allows key business indicators like products shipped or orders processed to be tied to IT resources.
BEZ monitors application and database tiers and support multiple databases, application services, operating systems etc. It allows customers to be more efficient through what-if analysis, automated notifications and easy change to configurations. Over time BEZ helps companies right-size their infrastructure and avoid service level problems. For instance, companies can see what they will need, in terms of hardware and software, to support a major promotion or product launch.
The BEZ system involves lots of data gathering, integration, and allocation to workloads for different business applications. Then BEZ predicts what might happen for different scenarios. Users can define service level agreements in advance or use predictions to set them. Business planning drives scenarios and growth numbers. BEZ allows the user to predict how proposed changes to capacity, tuning etc will affect applications’ performance so that choices/decisions can be justified in terms of business capability. Also support operational decisions that need to made right now in terms of load balancing, tuning etc.
Change is what drives BEZ adoption – companies that not only have a large and complex environment but one with change. Add in the fact that the ability to do live data warehouse workload management is fairly new and companies need help. While it is easy to make the changes, most IT departments don’t know what change to make and BEZ helps them understand it and map it to business needs.
All in all BEZ was an interesting use of analytics in the IT department. Hopefully their users realize that the whole company could also benefit from predictive analytics, once they see the obvious value to the IT department itself.
