By using website you agree to our use of cookies as described in our cookie policy. Learn More

TDWI Upside - Where Data Means Business

Q&A: AI’s Place in a Data-Driven Enterprise

AI is making BI and analytics more accessible beyond data analysts. MicroStrategy’s Saurabh Abhyankar, executive vice president and chief product officer, explains how.

Upside: Enterprises have been amassing “big data” for decades. What does a data-driven business look like in practice?

For Further Reading:

From Data-Driven to Data-Centric: The Next Evolution in Business Strategy

Three Areas Where AI Can Make a Huge Difference Without Significant Job Risk

Executive Q&A: Data Quality, Trust, and AI

Saurabh Abhvankar: Until recently, “data-driven” only applied to upper management because data dashboards were only available to them. Although it’s critical that C-level and VP-level executives make decisions based on relevant and accurate data, all employees make decisions that affect the business on a regular basis. In a truly data-driven business, everyone has easy, immediate access to the data they need to make accurate decisions quickly.

Why is building a data-driven enterprise so hard to do?

Mainly because people below the C-suite and the VP level only have sporadic access, at best, to data for decision-making. The breakneck pace of modern business is accelerating, and people rightly worry that they can’t wait for easily accessible and timely data before making a decision. Additionally, most people are not used to making data-driven decisions as part of their day-to-day work because they’ve never had fast and easy access to the data they need. Once everyone in the organization has easy access to data, making the transition to pervasive, data-driven decision-making requires a cultural shift and strong change management.

BI brought data analytics to senior leadership and the C-suite. Why is it challenging to use BI throughout the organization?

Providing access to everyone is difficult within a traditional BI platform because creating dashboards can be complex and time consuming so setting them up for every employee is not feasible. Even if you do set up the dashboards, they may require a steep learning curve or be too cumbersome to use effectively. Let’s say a sales rep is on a call with a prospect, and she needs to know whether a specific product could be delivered by a specific date. She’d have to leave the application in which she is working, find the right dashboard (assuming it’s already been set up), click through to the right data, and then return to the original app. That takes a lot of time, perhaps too much for a prospect in a hurry.

Is generative AI ready for prime time in a data-driven organization given its well-known problems with hallucinations?

Hallucinations are definitely an issue for generative AI, and that’s why an enterprise’s senior leadership cannot rely on it alone when they need information on which to base critical decisions. Executives need to have confidence that the data they’re looking at is reliable, and with generative AI alone, there’s significant risk being introduced.

Business intelligence (BI), on the other hand, produces data that’s extremely reliable, but it’s not always simple to access. That’s where generative AI can play a role. When generative AI is paired with BI with effective guardrails in place, all employees can interact with accurate BI data in a natural way using ordinary language and get results in any format they like: tables, summaries in natural language, graphs … the possibilities are limitless. It’s a perfect application for generative AI, because the bedrock accuracy of BI provides a solid foundation for getting accurate information, enabling organizations to get the benefits of generative AI without the risk of hallucinations.

What do those “effective guardrails” you mentioned look like?

Generative AI needs a variety of guardrails to ensure it doesn’t invent data points. If employees from C-level executives to frontline workers are going to make decisions based on generative AI responses, data governance is critical to ensure the data is reliable. Pairing generative AI with a trusted BI platform can provide that grounding in trusted data.

For Further Reading:

From Data-Driven to Data-Centric: The Next Evolution in Business Strategy

Three Areas Where AI Can Make a Huge Difference Without Significant Job Risk

Executive Q&A: Data Quality, Trust, and AI

There are broader issues as well that require additional guardrails, such as AI transparency (people need to understand how AI comes to its conclusions) and ethical use. The needs of each organization, however, will be different, so training that incorporates these guardrails will need to be customized for each organization.

Could you give a few examples of how the combination of BI and generative AI would play out in specific enterprise use cases?

The best examples are those where frontline or edge workers need to rely on data from multiple sources to make decisions under rapidly changing conditions. Retail merchandising is a good example. To maximize sales and optimize inventory every day, merchandisers use data from POS, ERP, retail operations, and many other systems.

Typically, these people are retail experts, not data analysts, and without extensive integration and analysis work done by IT behind the scenes, this data is very hard to correlate and use by the merchandiser for timely decisions. However, by integrating AI-powered BI and analytics into their workflows, retail merchandisers can use generative AI to ask questions about inventory levels, sales forecasts, supplier lead times, and historical data to optimize inventory replenishment. Not only does this save merchandisers time, it also reduces carrying costs and allows them to make data-driven decisions quickly, improving operational efficiency, enhancing customer satisfaction, and driving revenue growth in the highly competitive retail industry. This is exactly the type of thing our retail customers are doing with our Auto AI assistant integrated into our BI platform as well as their retail web and mobile applications.

What other technologies or trends are important to realizing the vision of a data-driven enterprise?

Digital transformation continues to be a massive market and critical for bringing legacy environments into the modern age of containerized cloud apps. However, digital transformation is extremely complex, expensive, and takes years to complete, so enterprises are also looking for ways to increase access to siloed data and applications and reduce the wasted time associated with jumping between dozens of apps to find the piece of information you need for a decision.

Hyperintelligence is a technology we developed to address this need. With hyperintelligence, organizations can integrate data and analytics directly into web apps with zero programming. It uses hypercards which pop up dynamically within an application to display data from any other system that is related to various keywords within the application the user is using, such as product ID, customer name, order number, etc. Users can then get real-time answers by hovering over these keywords in a web browser, scanning a barcode on a mobile device, or clicking on an email to see relevant data, in context and within their normal workflow.

Hyperintelligence is also integrated with AI to let users query the hypercard with natural language, digging deeper into the data to get more insights, or even triggering actions within another app, all without leaving the current application. It’s a huge game-changer in streamlining access to enterprise data and apps at a fraction of the cost and time that the typical data transformation project would require to achieve the same thing.

Once an enterprise realizes this dream of becoming data-driven at all levels, what benefits will it see as a result?

The biggest benefit is that people throughout the organization make decisions based not on their gut and intuition but on hard data, which means those decisions are more likely to be accurate. This transformation has an enormous impact on the entire company. McKinsey, for example, found that companies that implemented a data-driven sales growth strategy saw EBITDA increases of 15% to 25%. This study only looked at the sales organization. The impact will certainly be even greater when all employees have immediate access to relevant data insights.

[Editor’s note: Saurabh Abhyankar is the executive vice president and chief product officer at MicroStrategy. He has been innovating in the analytics market for 20 years and holds several patents in self-service analytics, the semantic graph, and hyperintelligence. Since 2016, he has held various product leadership positions at MicroStrategy including SVP of product management and EVP of marketing. Mr. Abhyankar received a B.Sc. in computer science from the University of British Columbia.]

TDWI Membership

Accelerate Your Projects,
and Your Career

TDWI Members have access to exclusive research reports, publications, communities and training.

Individual, Student, and Team memberships available.