Enterprises Grapple With AI Implementation Hurdles Amid Data Constraints

A few years back, the business technology landscape was dominated by the buzzword 'Big Data' – referring to the massive accumulation of information by organizations, which promised to uncover novel operational approaches and suggest optimal strategic directions.

What's increasingly clear today is that the challenges companies encountered in leveraging Big Data effectively persist, and it is the emergence of a new technology – AI – that is bringing these very issues back into focus. Without addressing the fundamental problems that plagued Big Data, AI implementations are destined to fall short.

So, what exactly is preventing AI from living up to its potential?

Most obstacles originate from the data resources themselves. To grasp the issue, consider the diverse sources of information utilized during a typical workday.

In a small-to-medium enterprise (SME):

• Spreadsheets saved on user laptops, in Google Sheets, or within Office 365 cloud storage.
• The customer relationship management (CRM) platform.
• Email correspondence among colleagues, customers, and suppliers.
• Word documents, PDF files, and online forms.
• Instant messaging applications.

In a large enterprise:

• All the above sources, plus,
• Enterprise resource planning (ERP) systems.
• Real-time data streams.
• Data lakes.
• Separate databases supporting various standalone applications.

It's important to note that this brief list is not exhaustive, nor is it meant to be. Its purpose is to show that even within just five bullet points, information can be scattered across roughly a dozen locations. What Big Data required – and still often does – and what AI initiatives also depend on, is a method to unify these disparate elements so that computer algorithms can interpret them effectively.

In Gartner's 2024 Hype Cycle for Artificial Intelligence, AI-Ready Data was positioned on the ascending slope, with an estimated two to five years before reaching the 'plateau of productivity'. Given that AI systems rely on mining and extracting data, the majority of organizations – excluding the very largest – lack the necessary foundational infrastructure. As a result, they may not benefit from AI-driven support for another one to four years.

The core challenge for AI deployment mirrors the one that hindered Big Data initiatives as they progressed through the hype cycle – from innovation trigger, peak of inflated expectations, trough of disillusionment, slope of enlightenment, to plateau of productivity. Data exists in multiple formats; it can be inconsistent; it may follow different standards; it might be inaccurate or biased; it could contain highly sensitive information, or be outdated and irrelevant.

The process of transforming data to make it AI-ready remains just as critical today – if not more so – than ever before. Companies aiming for an early advantage could experiment with the numerous data treatment platforms available. Following common recommendations, they might start with limited, discrete projects as test beds to evaluate the effectiveness of emerging technologies.

A key benefit of modern data preparation and assembly systems is that they are specifically engineered to organize an organization's information resources for use by AI value-creation platforms. For instance, they can incorporate carefully coded guardrails to help ensure data compliance and prevent users from accessing biased or commercially sensitive information.

Nevertheless, creating coherent, secure, and well-structured data resources continues to be an ongoing challenge. As organizations accumulate more data through daily operations, maintaining up-to-date and accessible data resources is a continuous effort. While Big Data could often be treated as a static asset, data intended for AI consumption must be prepared and processed in as close to real-time as possible.

Consequently, organizations must strike a delicate three-way balance between opportunity, risk, and cost. The selection of the right vendor or platform has never been more critical for the modern enterprise.

(Source: “Inside the business school” by Darien and Neil is licensed under CC BY-NC 2.0.)

Interested in learning more about AI and big data from top industry experts? Attend the AI & Big Data Expo in Amsterdam, California, and London. This comprehensive event is part of TechEx and co-located with other premier technology gatherings. Click here for additional details.

AI News is brought to you by TechForge Media. Discover more upcoming enterprise technology events and webinars here.