Dec 17, 2020

Beware of creating a legacy of artificial intelligence silos

Adam Mayer
5 min
Companies appear to be falling back into the silo trap by not distributing their artificial intelligence (AI) and machine learning (ML) capabilities
Companies appear to be falling back into the silo trap by not distributing their artificial intelligence (AI) and machine learning (ML) capabilities...

While the issue of silos in IT and data management are well-known, companies appear to be falling back into this trap by not distributing their artificial intelligence (AI) and machine learning (ML) capabilities across their business. New research from Qlik and IDC revealed that just 20% of businesses widely distribute these capabilities across the organisation.

However, with the rise of analytics solutions that leverage AI and ML to augment users’ experience of and insights from data, many business leaders are recognising that having these capabilities siloed in Business Intelligence teams will prevent them from generating the greatest value from their data. In fact, 42% consider expanding the use of AI and ML amongst workers as critical to improving the success of data analytics projects.

Three sources of silos

So, why are these silos arising once again? There are three key reasons, which many data leaders will be painfully familiar with.

The first is that many companies have gatekeepers to data across the organisation. This, in and of itself, is not an issue as this approach often provides the simplest option for governance by keeping data secure. However, it does limit the opportunity of certain areas of the business to take advantage of all the data they need to run advanced analytics tools that incorporate AI and ML to augment users’ intelligence. As such, there needs to be a better balance between meeting the needs of IT and the business.

The second challenge is that there are some sources where it is difficult to get the data out – or where if you don’t do it in the right way, the data isn’t as useful as it should be. ERP systems, like SAP, are a prime example of this and limit the ability for business functions, like Sales, to incorporate its data into intelligent analytics solutions for predictive modelling.  

Finally, many companies don’t have the skills widely dispersed across the organisation to support a more democratised use of AI and ML. Research from Qlik and Accenture previously revealed just 18% of employees globally report that everyone in their organisation has the skills they need to read, work, analyse and argue with data proficiently. Without these core data literacy skills, many knowledge workers will be unable to question and challenge the insights from intelligent solutions.

Democratising the benefits of AI and ML in data analysis

Understanding the issue is halfway to solving the problem. Those IT and data leaders that take affirmative steps now can break down these silos, so that their entire organisation has the potential to drive Active Intelligence from its data.

But, how can businesses successfully overcome the aforementioned challenges and increase the use of intelligent insights across their whole organisation?

  • Empower users to self-serve data - Given nearly two-thirds of business leaders (61%) cite that finding valuable data sources is one of their greatest challenges, the benefits of creating a searchable data catalog cannot be overstated. For example, a sales leader might search “customers” to be shown relevant data sets, from invoice to customer service data. Implementing it as a searchable SaaS platform rather than a static data store also supports the management of governance and access privileges. This provides a single, self-serve data catalog for a consistent user experience, which ensures people can only access the right data for their role.
  • Unlock the potential of raw data sources - ERP and CRM systems hold masses of valuable data, but providing near real-time access to this data in a format that is optimised for the read processes of analytical systems is a massive hurdle that prevents CIOs and CDOs putting it in the hands of business users. The traditional process of extract, transform, load (ETL) used to move this transactional data to data warehouses where it can be governed, cleansed and queried often takes between six to nine months, by which point much of its value might be lost to the business. Switching to ELT and automating the process of streaming data with Change Data Capture (CDC) enables organisations to access real-time information from ERP and CRM systems, in turn fuelling advanced and predictive analytics engines for business users.
  • Choose intuitive platforms – With a small fraction of knowledge workers capable of AI and ML analysis, organisations must choose augmented analytics platforms that significantly reduce the barrier to actionable insights. Intelligent systems can support users on their journey to finding the right information: for instance, conversational analytics help users intuitively navigate data, while natural language processing removes the barrier of technical language and centres on user intent. Procuring platforms for AI and ML analysis that require more specialist expertise will significantly reduce the accessibility for knowledge workers and establishes a significant hurdle for a decentralised approach.  
  • Invest in employee skill sets While our research with IDC revealed that currently just 16 percent of knowledge workers globally are equipped to do AI and ML analysis, it is encouraging to see that there are clear intentions to upskill more workers in this key area. Respondents predicted that this figure would rise to 25% of the workforce over the next two years, as well as increasing the proportion of those with data literacy skills from 45% to 63%. The role that employees’ skills play in removing barriers to data-informed decision making cannot be underestimated. These skills enable users to find, explore, analyse and question the key insights that AI and ML platforms generate, and which ultimately inform action and create positive business outcomes.

Beware of silos – again

Dismantling embedded, legacy silos continue to pose challenges to IT and data leaders the world over. As organisations embark towards a future of more intelligent analysis - with AI and ML enabling more proactive, personalised and collaborative experiences of data insights - these same leaders must ensure that they don’t fall into the trap of silos again. Democratising the benefits of augmented analytics will not only improve the experience and outcomes of organisations’ analytical projects today, but will lay the foundations for more lucrative insights that will drive truly Active Intelligence in the future.

By Adam Mayer, Senior Manager at Qlik

Share article

Jun 15, 2021

The advantages and disadvantages of AI in cloud computing

3 min
AI is being used in cloud computing, which works by allowing client devices to access data over the internet remotely, but are there pros and cons?

Cloud computing offers businesses more flexibility, agility, and cost savings by hosting data and applications in the cloud. AI capabilities are now combining with cloud computing and helping companies manage their data, look for patterns and insights in information, deliver customer experiences, and optimise workflows.

We take a look at some of the benefits and drawbacks of AI in cloud computing. 

The benefits of AI in cloud computing


Lower costs

A major advantage of cloud computing is that it eliminates costs related to on-site data centers, such as hardware and maintenance. Those upfront costs can be restrictive with AI projects, but with cloud enterprises you can access these tools for a monthly fee, making research and development related costs more manageable. AI tools can also gain insights from the data and analyse it without human intervention, reducing staff costs.

Deeper insights 

AI is able to identify patterns and trends in large data sets. Using historical data, AI compares it to the most recent data, which provides IT teams with well-informed, data-backed intelligence. AI tools can also perform data analysis fast so enterprises can rapidly and efficiently address customer queries and issues. The observations and valuable advice gained from AI capabilities result in quicker and more accurate results.

Improved data management

AI enables extensive data management, and cloud computing maximises information security, making it possible to deal with massive amounts of data in a programmed manner to analyse them properly, allowing the business to leverage information that has been “mined” and filtered to meet each need. AI can also be used to transfer data between on-premises and cloud environments. 

Intelligent automation 

Businesses use AI-driven cloud computing to be more efficient and insight-driven. AI can automate repetitive tasks to boost productivity, and also perform data analysis without any human intervention. IT teams can also use AI to manage and monitor core workflows. IT teams can focus more on strategic operations while AI performs the mundane tasks. 

Increased security 

With businesses deploying more applications in the cloud, security is crucial in order to keep data safe. IT teams can use different AI-powered network security tools which can track network traffic, they can flag issues, such as finding an anomaly. 

The drawbacks of AI in cloud computing


Data privacy 

 Enterprises need to create privacy policies and secure all data when using AI in cloud computing. AI applications require a large amount of data, which can include consumer and vendor information. While some data can be anonymous and can't be tied to personally identifiable information, knowing who the data belongs to makes it more valuable. When sensitive information is used, data protection and compliance is a major concern.

Connectivity concerns 

IT teams use the internet to send raw data to the cloud service and recover processed data. Poor internet access can hinder the advantages of cloud-based machine learning algorithms, as cloud-based machine learning systems need consistent internet connectivity. 

While processing data in the cloud is quicker than conventional computing, there is a time lag between transmitting data to the cloud and receiving responses. This is a significant issue when using machine learning algorithms for cloud servers, where prediction speed is one of the primary concerns.

Share article