05 Feb Business reporting has changed have you?
Business reporting has changed, have you?
Business Intelligence (BI) architectures today need to provide much more than the overnight canned reporting and dashboarding needs of the traditional system.
Here are four use cases:
Structured Reporting and Analysis
This use case doesn’t go away. Businesses still need regular reports and access to performance dashboards to support management. What is changing here is the reduced reliance on using reports to extract data sets into Excel for onward analysis, replacing this will ever more functional dashboard tools, and increased use of self-service data access and analysis.
Ever more functional dashboard tools, and increased use of self-service data access and analysis
Data scientists need access to raw source data in volume to perform their magic. They do need good quality data and can spend a lot of time cleaning data in preparation for their analyses. Some of their techniques (deep learning) don’t need to know schemas, but most do. A modern data warehouse can help by being a one stop shop for storing and serving data in volume, pre-cleaning some data (as it is used for other purposes) and maintaining schemas.
Logs are analysed for four purposes: to report system performance; to detect problems or indicators of future problems; to analyse what went wrong after a failure; and to find new ways of indicating problems. All these analyses need access to log data, usually stored as lots and lots of dated files, all formatted the same way.
Increasingly Business Intelligence is embedded inside critical business processes
BI used to be tacked on to production systems and used by management to monitor performance versus plan, and to conduct a few what if analyses. If the system went down it was non-critical, inconvenient yes, but the business could survive several days outage, or some loss of data. Not anymore. Increasingly BI is embedded inside critical business processes, it makes decisions and acts autonomously, feeds robots, informs decision-makers. Data needs to be fed much more frequently (or even real time), and made available for analysis as soon as possible, Downtime is no longer allowed.
Datavault specialises in developing architectures fit for today’s and the future’s needs. Come and talk to us.