Most understand the sage advice not to use technology for technology’s sake. But in the age of wow-factor digital technologies like advanced analytics, machine learning and artificial intelligence (AI), it can be hard to resist. In this Q&A, Dragan Rakovich, DXC Technology’s chief technology officer for analytics and a Distinguished Technologist, offers advice on how to focus the cool technology on valuable business outcomes.
Q: Why are analytics such an important component of any digital transformation?
A: Data analytics and machine learning help CIOs focus on increasing business value to their clients and delivering important business insights in the context of business processes.
Q: Is there a right way to get started?
A: Sometimes ad hoc analytics and machine learning start from, “Hey, this is really cool stuff. Look what we can do with the technology. Now where can we apply that in the business?” But it should be the other way around — put the business outcome first. When it comes to using data analytics and modern machine learning and AI, the key is identifying specific solutions with clearly identified business value to the organization. Start there.
Then you can test and create the algorithm. And then ask, “How do we bring this algorithm into the business process?” That should include cultural change, because it may require a completely different interaction between human beings and the business process that’s empowered by this kind of algorithm. On the other side, you can’t leave it unattended, because algorithms are not hard-coded application pieces. So if the data inputs change, the algorithm might behave differently and supply different outcomes to the business process.
That’s why you need to continuously monitor the algorithms. You have to capture the inputs and outputs in the algorithm every time it fires, so you can put them back into the dataset that’s used to train the algorithm and continuously keep improving the algorithm. This idea of continuous improvement, continuous change and continuous learning as the algorithm is implemented in the business process is very important.
Q: How far along are we in driving analytics into business operations?
A: Today most companies are doing quite a lot in analytics and machine learning and AI, but it’s still mostly ad hoc. There are a lot of proofs of concept proving the value. But very few organizations have actually implemented machine learning algorithms and analytics models directly into their business processes to improve their business operations. In fact, very few have even used analytics to change the business model of how they operate, based on augmenting the human portion of the process with machine learning. When it comes to machine learning and AI, this is still early days.
I believe many organizations are still missing a clear vision of how the business will leverage these new technologies. That’s one of the big challenges. Many CIOs lack a standardized approach to analytics and AI — commonly referred to as industrialized AI.
Q: Where’s the path to further progress?
A: CIOs need help with analytics. An AI platform could help them create the flow of information, both in batch and streaming of data end-to-end, with the algorithm as the end product. That would help them understand the behavior of the particular business process or the particular equipment they’re monitoring.
What companies need is not only a digital foundation for data analytics, but also an understanding of their long-term data analytics objectives. We can do that using multiple deployment options for public cloud, on-premises hardware, virtual private cloud or a hybrid environment.
Q: How is this playing out in the real world?
A: BMW Group has created a platform for research and development that is a key milestone in its path to autonomous vehicles — specifically, Levels 4 and 5 — that it hopes to have on the streets in the next few years. To give you an idea, BMW’s High Performance D3 platform, or data-driven development platform can store over 230 petabytes; the company gets some 1.5 terabytes of new data every day from its autonomous vehicles test fleet around the world.
For companies taking on autonomous driving development, they’re looking for a partner to create this type of technical solution and then manage it. Our accelerators and intellectual property enable the engineering, testing and data science people to deal with these increasing amounts of data in a rapid fashion. The platform can provide data for analysis in seconds rather than days or weeks. That data is ready for analysis, fully labeled and fully synchronized. That’s important, because you have all kinds of data coming from a car: video, GPS, sensory data and more.
Solving for the challenges of autonomous driving development is a great example of how technology can truly align with the business need to develop a completely new business solution. In autonomous driving there’s a rush to get to market first, so there is a clear business need.
To learn more, read the white paper, Use IT modernization to accelerate and scale business transformation.