Nov 3, 2021
Laura Berrill

When AI & data analytics meets business intelligence

Data
Analytics
businessintelligence
Rising from data analytics & AI, data and business intelligence is gaining popularity, so why is it important to today’s growth in data volumes?

A lot has been said about the value-creation opportunities an organisation can achieve from data analytics to unlock new insights into the customer experience. It can increase operational efficiencies and unleash new products and services based on a greater understanding of trends and untapped revenue. But, is this really the case?

We understand that data is driving new opportunities to transform business as a renewable resource, but more work still needs to be done.  The fact is data is growing. Figures released by IDC indicate that by 2025, the amount of data will double every 12 hours. This growth is set to continue with new forms of collection, such as natural language processing (NLP) which is beginning to have tangible effects.

Structural importance

However, a growing slice of the data pie is currently ‘unstructured’ and non-contextual. Consequently, this type of data has to be processed through the human brain. But now artificial intelligence and machine learning are gathering pace and becoming even more capable of analysing unstructured information.

Chris Stephenson, technical director at Sagacity explains: “AI and machine learning have great potential for driving the use of data analytics across a business, for example, identifying your most valuable customers, as well as those who are vulnerable and could require assistance.” 

However, he warns that you can’t hit the ground running straight away. “Before a business can successfully use AI for data analytics, it must first understand what data it has within the organisation and to which situations the data can be applied. Many organisations mistakenly think AI is a ‘plug and play’ technology that delivers returns straight out of the box, but you have to put a lot of work in before you see the results. AI and machine learning needs to be given reliable labelled training data before it can perform data analytics and draw valuable insights.”

So, in this respect, where does data intelligence come in and where does that leave the traditional data scientist, is it a death knell?

According to the CEO of Insights and Data at Capgemini, Zhiwei Jiang, AI data solutions being left to their own devices is ‘science fiction’. He says: “It’s an interesting question: Could data scientists be overthrown by new robots? The idea that AI solutions can be bought, plugged in and left to run on their own devices is, frankly, science fiction. AI and machine learning are doing more of the heavy lifting with regards to data science and analytics, but a deep affinity with algorithms and mathematical logic, even when the technology evolves into new areas (like deep learning and reinforcement learning), is vital. Data masters need to understand what lies beneath the surface if they’re going to make informed decisions about approaches and tools for the problems at hand.”

He advises that it will always be possible to explain data in human terms. “A data team which can clearly comprehend the complex metrics, maths and logic involved in any AI system is a no-brainer. Cold hard data means nothing if it can’t be clearly articulated. Emotional intelligence creates empathy, conversational capabilities and the ability to balance the objectives of being data-powered and being human,” he adds.

And Zandra Moore, CEO at Panintelligence, adds: “There will always be the need for people with the skills to analyse complex data and manage evolving data modelling requirements. As technology evolves, the volume of data and its complexity will continue to grow, so data scientists will continue to play a key role for businesses with the data management expertise they boast.”

With these developments we enter the realm of what we call ‘data intelligence’

With the exponential rise in data volumes across all industries, managed by both AI, machine learning and data scientists as described, data intelligence is the real fuel needed to accelerate past the competition in this field and create value. No company gets credit for simply being the best data collector around. It’s actually a risk and a liability if data is not curated and used responsibly.

So, what’s the secret? The best data-driven organisations have basically learned how to build the best data refineries. And that means increasing data intelligence for knowledge workers through self-service data analytics, rather than being stuck, or mired in the tar of unusable data lakes. These organisations have several things in common; they’ve taken all that raw data, curated it to understand what’s useful within their data lakes, catalogued it to make it fit for purpose and then democratised the data across the organisation to enable analytic insights from data intelligence. In essence, they’ve focused their data stewards on unleashing data intelligence to achieve actionable results based on metadata-driven insights.

Supercharged interest 

Rich Pugh, Chief Data Scientist at Mango Solutions, has more than 20 years’ experience working with data. For him, data and business intelligence will be driven to data AI solutions in its quest to keep in step with market competitors. 

“In recent years, the need to become a more intelligent, relevant and efficient organisation has given rise to significant investment in both data and advanced analytics,” he says. “We understand that, if data is the new raw ingredient, we need to dynamically turn this insight and wisdom to support decision-makers. This has supercharged interest in data science and AI and led to an increase in organisations looking to create data strategies that deliver (and sometimes define) their forward-looking business objectives. The broadening of the remit of data and analytics is also driving growth in data science teams. As a data scientist myself I can build the best model in the world, but if I can’t get someone in the business to change their behaviours to use the insight I’m generating, or if there are technical challenges that mean I can’t deploy my model in a repeatable way, then it stays as a beautifully crafted piece of code on a laptop.”

He went on: “The need to deploy data and analytic outputs has seen a significant increase in the need for data engineering teams who can build scalable and repeatable data pipelines and use DevOps approaches to put insight into the hands of the right decision-makers at the right time. For businesses to get the edge, hire great data scientists and arm them with the best data and business intelligence solution you can find for your market."

Share article