How generative AI is redefining data analytics 

The generative AI party is still raging. This zeitgeist has rocked the business world daily in a million ways, and the ground is still shifting. Now, four months into 2024, we’re starting to see businesses, particularly those with rarified pragmatic brands, starting to demand evidence of value, of the path to the true ROI derived from AI. As pragmatic voices for value rise, how do thoughtful business leaders respond?

Alteryx studied exactly this question. What are the concrete pathways to AI value? We surveyed leading CIOs and board members and found a brightly lit approach to engineering emerging AI capabilities into business outcomes.

Our survey found that generative AI is already impacting the achievement of organizational goals at 80% of organizations. What led the way, as the #2 and #3 use cases, were analytics—both the creation of and the synthesis of new insights for the organization. These use cases trailed only content generation in terms of embrace.

What makes analytics and generative AI such a potent combination? To explore that, let’s get started by diving into what key challenges generative AI solves for, how it works, where it can be applied to maximize the value of data and analytics, and why generative AI requires governance for success.

Overcoming analytics challenges with generative AI

Companies have long recognized the benefits of using data and analytics to improve revenue performance, manage costs, and mitigate risks. Yet achieving data-driven decision-making at scale often becomes a slow, painful, and ineffective exercise, due to three key challenges.

First, there aren’t enough experts in data science, AI, and analytics to deliver the breadth of insights needed across all aspects of business.

Second, enterprises are often hampered by legacy and siloed systems that make it impossible to know where data lives, how to access it, and how to work with it.

Third, even as we struggle with the first two challenges, data continues to grow in complexity and volume, making it much more difficult to use. Combined with a lack of robust governance policies, enterprises are then faced with poor data quality that can’t be trusted for decisions.

Applying generative AI to analytics

Generative AI presents two massive opportunities to tackle these challenges by improving the usability and efficacy of enterprise analytics tools.

Let’s talk about usability first. Generative AI makes analytics tools easier to use. Much of this is driven by the incorporation of natural language interfaces that make using analytics much easier, as the “coding language” can be simple natural language. It means that users can execute complicated analytics tasks using basic English (natural language) instead of learning Python. As we all know, coding languages have a high learning curve and can take years to truly master.

Next, in terms of efficacy, generative AI substantially improves the quality of automation that can be applied across the entire data analytics life cycle, from extract, load, and transform (ELT) to data preparation, analysis, and reporting.

When applied to analytics, generative AI:

• Streamlines the foundational data stages of ELT: Predictive algorithms are applied to optimize data extraction, intelligently organize data during loading, and transform data with automated schema recognition and normalization techniques.
• Accelerates data preparation through enrichment and data quality: AI algorithms predict and fill in missing values, identify and integrate external data sources to enrich the data set, while advanced pattern recognition and anomaly detection ensure data accuracy and consistency.
• Enhances analysis of data, such as geospatial and autoML: Mapping and spatial analysis through AI-generated models enable accurate interpretation of geographical data, while automated selection, tuning, and validation of machine learning models increase the efficiency and accuracy of predictive analytics.
• Elevates the final stage of analytics, reporting: Custom, generative AI-powered applications provide interactive data visualizations and analytics tailored to specific business needs. Meanwhile, natural language generation transforms data into narrative reports—data stories—that make insights accessible to a broader audience.

Top generative AI use cases for analytics 

The impact of generative AI for analytics is clear. Integrating generative AI in analytics can unleash the capabilities of large language models and help users analyze mountains of data to arrive at answers that drive business value. Beyond content generation, the top use cases for generative AI are analytics insight summary (43%), analytics insights generation (32%), code development (31%), and process documentation (27%). 

Alteryx is well-equipped to support a range of generative AI applications, including the following use cases, offering both the tools for development and the infrastructure for deployment: 

• Insight generation: Generative AI can work with different data sources and analyze them to provide insights for the user. To add more value, it can also provide and summarize these insights into more digestible formats, such as an email report or PowerPoint presentation.
• Data set creation: Sometimes, using real customer or patient data can be costly and risky but generative AI can create synthetic data to train models, specifically for heavily regulated industries. Using synthetic data to build proof of concepts can accelerate deployment, save time, and reduce costs—and even more importantly, reduce the risk of violating any potential privacy laws or regulations.
• Workflow summary and documentation: Generative AI can automatically document workflows to improve governance and auditability. 

Building a holistic, governed approach 

While there are endless opportunities for automation and new use cases that have yet to be discovered, leaders must understand that the trust of AI and LLMs is reliant on the quality of data inputs. Insights generated by AI models are only as good as the data they have access to. Generative AI success requires enforcing data governance in responsible AI policies and practices for AI adoption. 

On its own, using generative AI without guardrails can lead to data privacy concerns, inaccurate results, hallucinations, and many more risks, challenges, and limitations. It’s important for enterprises to work with vendors who have principles and frameworks in place that align with industry standards to ensure they can responsibly adopt generative AI at scale. 

To help enterprises mitigate these risks, Alteryx bakes in different mechanisms within its platform to control these challenges and simplify the AI governance process across the life cycle, while remaining grounded in principles that help us and our customers adopt AI responsibly.​ For example, we’ve built our platform to provide private data handling capabilities, allowing our customers to take their AI training and deployment entirely within their own firewall. 

Finally, it is critically important to implement proper controls and incorporate human-in-the-loop feedback mechanisms to enable ongoing verification and validation of AI models. This ensures their accuracy, reliability, and alignment with desired outcomes. 

Engineering emerging AI capabilities into business outcomes 

When used responsibly and in a secure, governed manner, generative AI can lead to key benefits such as market competitiveness (52%), improved security (49%), and enhanced product performance or functionality (45%). 

With the Alteryx AiDIN AI Engine for Enterprise Analytics, Alteryx makes navigating the generative AI landscape within an organization smoother and more manageable for analytics. Overall, the platform helps organizations get value from their generative AI investments by applying generative AI to their data to enhance customer experiences, streamline operations, and drive personalized interactions. 

Asa Whillock is vice president and general manager of machine learning and artificial intelligence at Alteryx.

—

Generative AI Insights provides a venue for technology leaders—including vendors and other outside contributors—to explore and discuss the challenges and opportunities of generative artificial intelligence. The selection is wide-ranging, from technology deep dives to case studies to expert opinion, but also subjective, based on our judgment of which topics and treatments will best serve InfoWorld’s technically sophisticated audience. InfoWorld does not accept marketing collateral for publication and reserves the right to edit all contributed content. Contact doug_dineley@foundryco.com.