How AI is shaping the future of business intelligence

Business imperatives for modern BI

The adoption of modern BI is driven by clear business needs. In today’s dynamic environment, leaders can no longer afford to rely on gut instinct. “Instead, they need access to timely and trusted data insights to make critical decisions with confidence,” says Nate Nichols, vice-president of product at Tableau.

Organisations are also turning to BI to build resilience and agility in the face of uncertainty. “Whether it’s optimising operations, improving customer experiences, or navigating supply chain volatility, analytics help them respond more quickly and confidently,” notes Garavello.

Furthermore, as businesses constantly look for ways to do more with less, BI can pinpoint inefficiencies and bottlenecks to improve operational efficiency and reduce costs, adds Luca Spinelli, managing director for ASEAN at SAS.

This has translated into real-world results. CIMB Singapore, a consumer bank in ASEAN, adopted the SAS Viya platform to gain a consolidated view of its customers. This foundational step has not only unlocked insights into the bank’s customer segments but also reduced the amount of time its employees spend on finding the right data, from 80% to just 20%.

Traditional dashboards gave us hindsight. Modern BI provides us with foresight, context, and, increasingly, autonomy. We’re moving from dashboards to decisions

In Japan, Nissin Foods Holdings modernised its data environment with Qlik to transform inventory and demand planning, enabling real-time decisions and supporting a data-driven culture. In India, IBM’s work with the State Bank of India has reduced report generation time from days to minutes, providing critical real-time operational insights.

To achieve these goals, organisations often leverage a spectrum of analytics capabilities, reflecting different stages on the data maturity curve. Each stage provides increasing value, moving from description to prescription. Patrick Kelly, senior director for Southeast Asia and Greater China at Databricks, outlines the journey: 

• Descriptive analytics (What happened?): The foundation, providing visibility into historical performance, helps spot trends over time, and ensures transparency in operations.
• Diagnostic analytics (Why did it happen?): Digging deeper to understand root causes using techniques like drill-downs, correlation analysis and data mining. It helps teams understand the factors influencing outcomes so they can respond more effectively, whether fixing issues or building on what works.
• Predictive analytics (What is likely to happen?): Using historical data and AI to forecast what might happen next, such as changes in customer demand, sales performance or potential risks. It helps teams understand the factors influencing outcomes, so they can respond more effectively, whether fixing issues or building on what works.
• Prescriptive analytics (What should we do about it?): The most advanced form, recommending specific actions to achieve desired goals using simulations or optimisation algorithms. It helps organisations make smarter, data-backed decisions, maximising gains, reducing risks and even automating complex choices at scale.

The data paradox

Despite the clear benefits and advanced tools, the path to BI maturity is fraught with challenges. Many organisations find themselves in what Tableau’s Nichols calls a data paradox. “They need to make data-driven decisions faster than ever before but are hampered by large volumes of often unreliable data spread across different sources,” he says. “The result? Businesses are data-rich but insight-poor.”

This paradox is rooted in several interconnected issues:

1. Data silos and integration complexity

Data is fragmented across on-premises legacy systems, multiple cloud platforms and countless cloud-based applications. Integrating this disparate data is a monumental task.

“To address these challenges, leading companies are taking a more unified approach to their data infrastructure,” says Databricks’ Kelly.

Tools such as Databricks’ Lakeflow Connect for data ingestion and Tableau’s Zero Copy Partner Network, which allows querying data at its source without moving it, are designed to tackle this complexity head-on.

2. Data quality and governance

The most sophisticated AI model is useless if it’s fed bad data, with Garavello adding: “Trust is everything in analytics, and it starts with data quality and governance.” This requires a concerted effort, often led by a chief data officer (CDO), to establish clear stewardship, data pipelines and governance frameworks.

Anup Kumar, distinguished engineer and head of client engineering for IBM Asia-Pacific, notes that data quality requires ongoing attention and continuous improvement, not a one-time fix. “The steady and sustained focus that a CDO organisation brings delivers results that are both more consistent and progressive over time,” he says.

3. User adoption and a data-driven culture

A new BI platform is only valuable if people use it, and adoption falters without trust and usability. “The biggest barrier to BI success goes beyond technology – it’s about earning trust, ensuring usability, and driving alignment,” says Garavello.

Kumar agrees, noting that “trust in the data remains the fundamental key to achieving widespread adoption”. Cultural resistance, a lack of data literacy, and a failure to demonstrate clear value to employees in their specific roles are common roadblocks. As SAS’s Spinelli observes, users might see a new tool as “more work” or an “IT colleague’s job” if its benefits aren’t immediately apparent.

From analysis to conversation

The solution to many of these challenges, particularly user adoption and complexity, lies in artificial intelligence. “AI isn’t replacing BI but redefining the experience. We’re moving from dashboards to co-pilots, from reactive analysis to proactive nudges,” says Garavello.

The growing use of AI in analytics is emerging under various names – augmented BI, agentic analytics, conversational BI – but the goal is the same: to make data analysis as simple as having a conversation.

Both Nichols and Garavello noted the rise of agentic analytics, where users collaborate with AI agents to automate and streamline analysis. Databricks has put this into practice with its AI/BI Genie, which Kelly describes as a tool that lets anyone “talk with their data”. Users can ask natural language questions like, “Why did sales spike in April?”, and receive instant, governed answers without writing a single line of code.

By lowering technical barriers, simplifying access and aligning with how business users work, such tools are helping organisations overcome adoption challenges and foster a culture of data-driven decision-making.

For example, NTT Docomo, one of Japan’s largest telcos, uses Databricks to analyse its usage of large language models (LLMs), reducing manual analysis time by 90%. At the same time, it has empowered broader teams with AI/BI Genie to drive natural language insights, spurring employee innovation and paving the way for generative AI adoption across its business.

Evolving skills

As AI automates the technical heavy lifting of data preparation and analysis, the skills required for a career in BI are evolving from pure technical proficiency towards a blend of business acumen, critical thinking and communication.

Garavello says: “We used to ask, ‘Can you code?’, and now we ask, ‘Can you interpret, explain and challenge what the AI tells you?’. You don’t need to be a data scientist to be great at analytics – what matters more today is being able to read data, question it and turn it into a good story.”

Nichols of Tableau echoes this, noting that as tedious work such as data cleaning and data crunching is automated, analysts can spend more time on strategic tasks. “This includes asking the right questions of the data, reviewing analysis and insights from AI and making sure it aligns with the business context, and tackling novel problems that aren’t easily automated,” he says.

They need to make data-driven decisions faster than ever before but are hampered by large volumes of often unreliable data spread across different sources. The result? Businesses are data-rich but insight-poor

The future belongs to what Garavello calls “data-native thinkers” who combine domain knowledge with curiosity. IBM’s Kumar adds that these professionals will need to deeply understand specific business domains – such as finance, healthcare, manufacturing and retail – and translate business problems into analytical questions that AI-powered systems can address. They will need to understand what questions to ask, how to interpret results in business context and how to act on insights effectively.

For businesses looking to enhance their BI and analytics capabilities, the advice from experts is clear and consistent. First, focus on the foundation.

“Don’t start with technology. Start with clarity,” says Qlik’s Garavello. Identify the key business decisions that need improvement and work backwards. This must be coupled with a robust data strategy.

There’s also a need to modernise the data fabric, the foundational infrastructure that determines the quality, accessibility and reliability of all analytics initiatives, says IBM’s Kumar. “Without a solid data fabric that can effectively integrate, govern and deliver trustworthy data across the organisation, even the most advanced analytics tools will struggle to deliver meaningful value,” he adds.

Businesses should also define what data maturity looks like and what it means to their organisations, Nichols says, adding that this begins with benchmarking competency levels and capabilities across their people, processes and technology.

From there, success can be measured through clear metrics, such as business performance, analytics productivity, organisational alignment, community satisfaction and adoption. “A data-mature organisation is one where data is accessible, trusted and actively used at all levels to inform decision-making throughout the organisation,” he adds.

SAS’s Spinelli says organisations will also have to invest deeply in upskilling employees at all levels. This goes beyond training the data team. Provide continuous, role-specific training, foster a culture where asking data-driven questions is encouraged and celebrate data-driven successes. This helps to overcome resistance to change and truly embeds a data-driven culture.

Ultimately, the future of BI is about creating a symbiotic relationship between human and machine. By combining AI’s power to process vast amounts of data with human curiosity, domain expertise and storytelling ability, organisations can move beyond simply understanding the past to actively and intelligently shaping their future. In a world facing a deluge of data, as Garavello says, “meaning is the new superpower”.