TDWI | Training & Research | Business Intelligence, Analytics, Big Data, Data Warehousing

Organizations Must Be Prudent To Realize Value In Generative AI

With all the new technology flooding the market, how do you know where to make the wisest AI investments? Here are three principles to guide you.

• By Rakesh Jayaprakash
• January 12, 2024

In today's fast-paced, AI-driven tech landscape, where a multitude of new technologies constantly emerge, your organization must exercise caution and prudence when choosing where to invest their resources. This diligence is especially crucial in the current economic climate marked by uncertainty and sluggish growth. Enterprises need to differentiate between flashy features and features that offer enduring value to empower businesses to make informed decisions and thrive in uncertain times.

For Further Reading: How Generative AI Is Changing How We Think About Analytics The Problem and Promise of Generative AI Generative AI Poses Security Risks

Look for Value in Capabilities Enabled by Generative AI

Many vendors are quick to incorporate generative AI into their existing products. Some of these implementations undoubtedly bring tangible value to end users, but it is equally clear that the majority merely add a superficial layer of generative AI without making a substantial impact, so your organization must exercise caution and discretion in your purchasing decisions when it comes to the technology. Rather than being swayed by the allure of generative AI capabilities, remain steadfast about the core features that can genuinely transform and enhance your operations.

This pragmatic approach should be considered a short- to mid-term strategy for any forward-thinking organization. The reality is that features closely coupled with generative AI capabilities are still on the horizon. It will be at least a couple of years before they become commonplace. To navigate this transformative landscape effectively as an analytics professional, you must equip yourself with a deep understanding of generative AI. This proficiency will enable you to distinguish between features loosely coupled with generative AI and features that are natively and seamlessly integrated into the technology stack.

Furthermore, keep a vigilant eye on the vendors supplying your critical business software. A vendor's stance and commitment to generative AI can profoundly impact how your organization operates in the future. For instance, the growing recognition of how large language models (LLMs) can benefit end users has prompted vendors to embark on their own LLM development, reducing their dependence on entities such as OpenAI.

However, it is important to acknowledge that such LLM initiatives may currently be on the early side of the maturity curve. Although there is no denying that generative AI is going to shape how we interact with software in the future, stick to the fundamentals in the short term rather than going after vendors betting big on generative AI.

Watch Out for the AIOps Maturity Curve

The current landscape of AIOps is largely characterized by vendors focusing on the reduction of alarm noise through automating the identification and correlation of recurring or similar alarms originating from various devices. Although this application of AIOps is highly specialized, the core principles behind this correlation process hold the potential for broader applications across diverse domains, enabling the recognition of patterns.

In the day-to-day operations of businesses, a multitude of events unfold across departments. By correlating these disparate events within the context of ensuring seamless, efficient operations, intriguing patterns can emerge. An example of this potential is identifying the root cause of security breaches in an organization. Various data points, such as endpoint vulnerabilities, compromised privileged accounts, anomalous login behavior, patch compliance, and unauthorized physical access must be analyzed to fully understand the root cause and impact of a security breach.

Such correlations necessitate the examination of data points spanning various data sets. Traditionally, connecting these data points required the expertise of data scientists. However, with the emergence of domain-agnostic AIOps, facilitated by automated ML capabilities that can integrate data from various domains, your organization can explore different avenues to enhance your operational efficiency and profitability. Although fine-tuning the underlying algorithms remains essential to cater to your organization's unique needs, the ability to uncover and leverage patterns can serve as a substantial competitive advantage.

Analytics Tools will Evolve into Automation Hubs

Analytics applications serve as the nucleus of an organization, providing a unified, comprehensive view by consolidating critical metrics essential for informed decision-making. Analytics applications are poised to transform into automation hubs in the coming years, with key elements such as organizational data and insights converging within these platforms to create the ideal environment for decision makers to implement automation. Organizations at lower levels of analytics maturity can initiate automation through preconfigured thresholds (the conventional approach) and subsequently progress towards automation triggered by ML algorithms that detect anomalies.

The pivotal advantage of initiating automation from analytics platforms will lie in overcoming the limitations associated with specialized software confined to specific domains. Unlike such software that is restricted by the type of automation it offers, analytics platforms will open up a multitude of possibilities for triggering automation. Your businesses will no longer be constrained by the automation capabilities of individual software; you will be able to harness the advantages of centralized automation. Given that most analytics applications already function as versatile platforms catering to various domains, the potential for automation is boundless.

You can prepare for this paradigm shift by ensuring that crucial data sources feed essential metrics into your centralized analytics tools. Additionally, employ ML algorithms to gain insights into data sets and behavioral patterns. This proactive approach will position your organization on a trajectory towards centralizing automation, thereby enhancing operational efficiency and agility.