In today’s business and IT world, data is the very sea we swim in, informing everything from purchase decisions to marketing and legal issues. It is also a reality that has hitherto given the advantage to large companies, as they are the only ones capable of meaningfully processing it.
New ways of looking at data, however, not to mention new strategies for dealing with it, may just change that.
New reality
Timandra Harkness, author of the new book ‘Big Data: Does Size Matter?’ says that the new reality is already upon us.
“Data is sometimes described as the new oil, but it’s much more than that. Increasingly, data is the way we relate to other people, whether we’re trying to hire them, date them, or win their votes,” she said.
This is, of course, where privacy concerns come in—something that is being addressed by the EU’s general data protection regulation (GDPR). The GDPR, which has been looming over business for two years, is already law across the EU, and will be enforced from 25 May, 2018.
The key to not only the GDPR but to understanding data today, is that data is not mere lumps of congealed information but something entirely new. Big data is far more than just masses of information, or systems to link datasets and analyse the—it is a new way of seeing the world, and each other. And that’s not necessarily a good thing, Timandra Harkness, author and broadcaster
In fact, says Harkness, a BBC radio science presenter, the key to not only the GDPR but to understanding data today, is that data is not mere lumps of congealed information but something entirely new.
“Big data is far more than just masses of information, or systems to link datasets and analyse the—it is a new way of seeing the world, and each other. And that’s not necessarily a good thing,” she told Techpro.
Businesses, however, have little choice but to get to grips with data—and, needless to say, to do so in a GDPR-compliant manner.
Taking advantage
Internet giants such as Amazon has built data processing—and retention—into their business models, but it is not only retail that can take advantage of data. However, without Amazon-like abilities to collect, store and process data, surely no-one can take advantage of it at all?
Oracle, as the world’s leading database vendor, has been at the centre of data-centric computing since its foundation. Of course, how we understand the meaning of the term ‘data’ has changed tremendously since then, not least because of widespread connectivity and the availability of previously unthinkable amounts of storage and computational power.
John Abel, Oracle’s vice-president of cloud and technology in UK, Ireland and Israel, says that data is particularly important in what he calls ‘the customer challenge’.
Customer challenge
“If you look at a [typical] retailer, for example, they’re trying to work out how do they take on the Amazons of the world. What we do know is that the Amazons have a couple of things in their favour and things against. In their favour, they have the ability to scale, both the expertise and the technology,” he said.
How, then, to challenge that? With data-as-a-service, he says.
What we know about AI is that the more data you give it, the more will you will get an improvement in probability. If you have IoT in your strategy and you connect it to your data store, it means you get AI up and working faster: the better the pattern searching and the better the results; there is an intrinsic connection to data volume, John Abel, Oracle
“The nice thing about Oracle is, we’re quite unique in the world in that we’re not competing with our customers. We’re nor becoming a bank, we’re not becoming a retailer. We’re a friend and can give the muscle to scale against the scaling companies.
“What that doesn’t get you to, though, is the concept of smart data, which is the ability to use technologies like AI [artificial intelligence] and ML [machine learning] to work with the data in a smart way,” he said.
AI and ML level the playing field, he says, because their deployment means data can be transformed into useful information without costly human intervention.
“Not everyone can afford the mass of data scientists—or the security people or the GDPR people,” he said.
Noise to signal
Oracle sees its data services as a means of systematically removing noise so as to enable businesses to make the right decisions.
Data should be, by its very nature, big, says Abel, as this is how behaviours and patterns can be identified.
“What we know about AI is that the more data you give it, the more will you will get an improvement in probability. If you have IoT in your strategy and you connect it to your data store, it means you get AI up and working faster: the better the pattern searching and the better the results; there is an intrinsic connection to data volume,” he said.
Being able to deal with data on a service basis not only brings businesses into line with the web giants, says Able, but could also be transformative in allowing them to focus on their core competencies rather than building up new cost centres. This will, he argues, eventually lead to the end of jobs and tasks based on ‘industrial logic’.
What we’re seeing now is a mixed approach where you’d have had analytics teams within the organisation as well as work with us. We have some customers looking for a partnership approach while others prefer to do it in-house, but, in any case, what we’re doing is trying to make it as easy as humanly possible to scale up, scale down, pull in your datasets and so on, Jason Burns, IBM
Machine learning
“Machine learning is the one I’m fascinated by, more than AI. You’re now having a machine-to-machine conversation. This is my personal view, not an Oracle view, but in any company going forward the thing you have to focus on is being creative in whatever area you’re in. Be creative and let the technology companies like Oracle, who aren’t going to compete with you, deal with the tech.
“Roles focused on industrialised logic will go to machine learning, but the human gift of creativity is needed—and that’s what AI and ML cannot do,” he said.
Data bet
Another company closely associated with big data is IBM: not only is it arguably the world’s oldest information processing outfit, it is also a company that has bet on data—in particular through the commercial deployment of its Watson artificial intelligence system—as being key to the future of business.
Despite its reputation as a service and solutions provider to massive, global conglomerates, IBM in fact has something for businesses of all sizes.
Jason Burns, IBM’s Ireland analytics architect and GDPR leader, says that this is very much company policy.
“In terms of size, we’re targeting everyone. That’s the nice thing about our cloud based environment; you can scale up on a pay-as-you-go model,” he said.
Unsurprisingly, however, certain industries and sectors are at the forefront of the change—in this case one particular industry that is no stranger to Irish shores.
“The customers I’m dealing with tend to be in the financial services industry, as they tend to be at the head of things.”
Data cleansing
According to Burns, any business that wants to get to grips with data must do so by considering the value of the data rather than concentrating on the technology. A key part of this is data cleansing, to ensure that it is actually readable and useful.
What I came to see more and more is that most businesses in Ireland don’t have a good understanding of data. In fact, one of the good sides of GDPR is that it may mean we see a better understanding of data, Eamon Moore, EMIT
“The whole thing is to not get vendor lock-in. A lot of the banks are trying to keep a lot of plates spinning: you spend a lot of time preparing data before you get to do the clever analytics. A lot of data scientists don’t want to be down in the bowels trying to get things to talk to one another.”
How businesses approach this will depend on what they want to get from data, and the nature of their overall IT strategy.
“What we’re seeing now is a mixed approach where you’d have had analytics teams within the organisation as well as work with us. We have some customers looking for a partnership approach while others prefer to do it in-house, but, in any case, what we’re doing at IBM is trying to make it as easy as humanly possible to scale up, scale down, pull in your datasets and so on,” he said.
Burns says that it is essential to keep a level head: data can bring huge benefits to business, including when it is managed via machine learning strategies that remove a lot of the drudgery, but there is already a lot of hype and pie-in-the-sky promises. something that can be made worse by fuzzy goals and bad planning.
Massive potential
“There’s the usual hyperbole that goes with a lot of these things. At the end of a day there’s just a file store. In terms of the impact on business in Ireland, there’s still massive potential. It’s not a mature business area, yet. It’s not just people developing stuff in Python; there’s still a huge demand for SQL skills as so much of the data is in relational databases.
“What we’re doing is we have a massive investment in AI. We’re trying to package our AI and ML into something that is useful. You can see that in Watson data services,” he said.
Eamon Moore, founder and managing director of Dublin-based solutions provider EMIT, says that data needs to be conceptualised properly in order to be of use—and the first step is to ensure it is the right data.
“Once the good data is there you can extract value. The data strategy is only as good the data being collected,” he said.
This may mean extensive data cleaning operations.
Data quality
“People put in phone numbers as, say, 087 or +353 87, or in some case, 00 353 87, and we see systems only looking certain strings. That’s obviously not good enough, so you want to build in some intelligence there.
“The less cleaning you have to do the better,” he said.
Nonetheless, the value is there to be extracted, he said—and Emit has committed to a business model of helping make sense of data, though, at present, the Irish business sector is lagging behind some other countries.
“When I look at the data-as-a-service model, two years ago we built out a new business unit at Emit.
“One thing we help companies do is around Power BI, the business intelligence suite from Microsoft. What I came to see more and more is that most businesses in Ireland don’t have a good understanding of data. In fact, one of the good sides of GDPR is that it may mean we see a better understanding of data,” he said.
Emit uses already well-known software tools at the front end to enable even the smallest of businesses to get to grips with the information currently locked up in their data.
Unifying silos
“Power BI can work on top of ‘siloed’ information and bring it together, creating dashboards around things like KPIs [key performance indictors]. I want to move people away from opening up e-mail first thing in the morning, toward opening up the dashboard to look at those KPIs.
“The as-a-service model means we can build in new things and bring them the reports that they need. Any BI solution out there will sit on top of data and give you insights, but the challenge we see is that of making sense of the databases. With legacy systems there is often no documentation and no understanding of the schema. That’s half the battle, frankly.”
Data-as-a-service does mean, however, that the needs of individual businesses can be catered for, and without significant capital expenditure on servers and software.
“It’s a consultation process. I say, don’t be led by the KPIs the software vendors provide, be led by the KPIs you want to report on. That’s why we provide a whole dataset on things many businesses didn’t know the compile reports on.
“We had one client in retail who wanted reporting from PoS [point of sale], and once they had that they had a better understanding of customers and were better able to market to them. They’ve seen a real increase in inline sales and in local sales.
“The question about getting into data, whether as a service or otherwise is: is there a way of getting value from the data? Well, yes,” said Moore.
Subscribers 0
Fans 0
Followers 0
Followers