The explosion of data in the recent years has had some knock-on effects. For example:
- Data theft is far more prevalent and profitable now than ever before. Ever heard of Crime-as-a-Service?
- There is now more pressure than ever before to modernize our applications to take advantage of the latest advances in DevOps and Cloud capabilities.
The consequence of this is that Dev/Test shops have moved towards eliminating the 'bad guy' by using subsetted or pure synthetic data throughout their SDLC. After all, it kills two birds with one stone: Data is small and easy to get it when they need it, and nothing of value exists to be stolen.
But the implication of this well-meaning act is that application quality decreases and their application projects are just as slow, if not slower, than before. Their use of non-realistic datasets results in an increase in data-related defects. Then they try to combat the self-inflicted quality issues by creating a whole new data program lifecycle around coverage mapping, corner cases, data quality, data release, etc. The net result is that they spend at least as much human and calendar time on data, as they did before...yet they will still have self-inflicted data-related quality issues.
We need to stop the madness. Data is not the enemy, rather it is the lifeblood of our companies. The true enemy is the same enemy we have been tackling with DevOps: Tradition. The traditional way that we have been dealing with the culture, process, and technology around data is the enemy. At Delphix we help our customers quickly flip this on its head and eliminate the true enemy of their business. By enabling our customers to provision full masked copies of data in minutes via self-service automation, they now have data that moves at the speed of business. Their applications release over 10X faster, their data-related defects plummet, and their surface area of data-risk decreases by 80%. And one of the beautiful things is that, in most cases, Delphix is delivering value back to the business inside of two weeks.