Data management is a complex thing. Ask anyone undergoing a digital transformation project and you will find that while they want to make the best use possible of their data, knowing where to start can be a serious challenge.
For Horia Selegean, data governance director at BT, the answer to this puzzle was governance.
Speaking to IT Pro at Hitachi Vantara's Next 2019 conference in Las Vegas, Selegean says: "BT is responsible for delivering it managed services for most of the larger multinationals across the world.
"That brings different complexity, the data brings different scales of data, and governance is operating 180 jurisdictions, you're going to have to comply with 180 local laws, and we have a lot of data that we manage. It is a very challenging and complicated environment, but really, really exciting in terms of what we can do with the data in the future. So it deserves a lot of focus and attention."
Indeed, for BT the drive to transform the way it analyses and consumes data from a siloed approach into more of a DataOps model started not with the company's data scientists pushing for a change, but from the introduction of GDPR.
"GDPR really give us a boost in our understanding and context around our data," he told delegates earlier in the day in the opening keynote. "We spent a lot of time landscaping, cataloguing, and understanding where the data is. That really helped us transform our function into a much more data-driven and governance-driven analytics [one]."
Expanding on this, he tells IT Pro that following the demands of GDPR "helps you to do things in a safe way. It allows you to protect data better, you put a better wrap around it, you've got a very good understanding of what data you have and where you have it".
Starting with governance, rather than having in mind a specific Big Data analytics project, may seem to be a more complicated way of working. But Selegean says it's quite the opposite.
"I would say starting with governance makes it easier the reason is, we operate for very, very large organisations, and those organisations that have specific needs, and they have specific demands from us," he explains. "Knowing and really respecting those demands is very, very important. And, to me, it's at the heart of everything we do in terms of how we safeguard the data and secure the data and look after it.
"Putting that into context will then allow us to operate, our analytics functions in a much more agile and much more specific way."
He doesn't deny that this may not be the speediest way to start using a DataOps strategy, but says: "Because the amount of time we spend at the beginning of the process to understand the data landscape, understand the needs, it pays back in terms of agility at the end of the throughout the life of the project. So yes, you may be slightly slower at the beginning to really set it up. But it actually pays off in the long term."
