Even as analytics emerged over the last few years as a core capability for businesses with a data driven decision making culture, companies often found themselves struggling to get the information they need out of disparate data silos.
Those that have done so, or made significant progress to that end, now find themselves in the envious position of being able to apply their data assets to transformational technologies like AI and machine learning, and automation. The laggards risk getting left behind.
Whether that is for mass personalisation, asset intelligence or internet of things implementations, all of these functions require data and integrations that need to form part of the AI and automation roadmap.
According to Darren Cockerell, Head of Solutions Consulting ANZ, Blue Prism, “When it comes to being able to harness AI technologies, the ability to manage data is everything. Digital disruption relies on the ability to ingest from, and disseminate to, legacy operations.”
Cockerall says there are already many impressive cognitive technologies available today, including those for marketers designed to gather insights to better understand and redefine the customer journeys “But every single one of these solutions follows the same paradigm; you have to get data in and then find a way to act upon what the solution delivers.”
Getting access to the cognitive service is the easy part since so many exist as cloud-based SaaS applications, says Cockerall. However, he cautions that corralling the data to feed to that service is the difficult part.
“The biggest barrier tends to be the volume of disparate legacy systems, spreadsheets and PDF documents across siloed departments. The data required is often voluminous and dispersed,” he says.
The impediments organisations face getting their data story straight are myriad, says Simon Belousoff, executive director of Beta Evolution, an independent digital, data analytics, and customer experience consultancy, and who was previously Head of Personalisation/ Customer Decisioning (Customer Transformation) at Bupa.
He says organisations often adopt a mindset and approach for data, CX and AI that is based on their legacy approaches to reporting. “What they really need is a different and evolved perspective and approach. This mistake often results in the data not being available in a timely way where it needs to be used.”
Unlike in previous processes, humans are often not directly involved
Furthermore, he says, “Data available for AI is consumed machine-to-machine at scale and needs to be consumable like this.”
He also cautions that operational silos are as corrosive as technical ones,
“Data is not seen as an enterprise asset, that is usable for the collective benefit of customers and the business. Instead it is seen as a discrete channel or function, or a business asset that is not for sharing with others in the organisation. You need to democratise the data.”
According to Belousoff, “Internal organisation data benefits from being progressively augmented with many forms of external data to deliver use case and experience outcomes and that this needs to be done in an integrated, timely and governed manner.”
Belousoff nominates the CBA’s Customer Engagement Engine which is powered by Pega and which saw 200 machine learning models created by Pega's AI based on the CBA's data scientist developed predictive models.
In a video describing the impact of that project, Angus Sullivan, Group Executive, Retail Banking Services, Commonwealth Bank of Australia said, “We just did our 50 millionth next best conversation in person."
That is 50 million times our frontline – either our branch or our contact center staff – had an opportunity to take the next best conversation and deliver that to our customers.”
“In digital, the number is multiples of that, “ he said, because every little interaction, it could be an email we send, or an alert that we pop up in the app, there the number is in the billions.”
Often technologies like AI and automation work hand in hand and practitioners say it is important to think holistically about these technologies rather than viewing them as discrete ideas.
According to Matt Oostveen, chief technology officer and vice president, Asia Pacific, and Japan, Pure Storage, “Not only is there a massive amount of data in existence but it is growing exponentially.”
Add to this the fact organisational data creation spans from the edge to the data centre and cloud, a surge in the amount of IoT connected devices, a need for decisions to be made in real-time via analytics workloads and we have reached a point where human operators cannot keep up with the digital tsunami, he says.
“Without automation, not only do we risk being swamped by data, but we may also miss the tremendous opportunities presented by AI and ML. Automation provides the platform upon which AI stands,” Oostveen says.