GenAI technology is set to revolutionize business, but its greatest impact will be on organizations, not just technology. The real shift is that businesses will delegate decision-making authority to AI in ways they’ve never done before.

In a recent fireside chat, George Fraser, CEO of Fivetran, sat down with Steve Jones, EVP of Data-Driven Business at Capgemini, and Bob Muglia, former CEO of Snowflake, to explore key strategies for AI preparedness, from data accuracy and governance to knowledge graphs and large language models (LLMs). Together, they dove into implementing strategies around data accuracy, data governance and LLMs, while also discussing the current and future state of GenAI and its reliance on clean, error-free data. 

“Technology is going to keep evolving. Every week there's a new model. Every day there's a new variation and a new piece,” Jones says. Those looking to succeed will need to take a thorough look at their data to ensure data sources, destinations and processes are where they need to be for GenAI to enter the picture. 

Data curation: The backbone of AI success

A key theme in the discussion was the critical role of data curation. While advanced algorithms capture headlines, the real work of AI often lies in managing the data that feeds these models. 

Despite the buzz around AI, many businesses are still in the proof-of-concept stage. While early experiments show promise, transitioning them into production presents a major challenge. “The gap between a good demo and a production-ready AI system has never been bigger,” said Jones.

To avoid this hurdle, businesses must first ensure the quality and availability of their data. AI can’t perform effectively without accurate, reliable information. The organizations that will succeed with AI are those that prioritize data quality and build a solid foundation to support AI’s growing role in decision-making. 

“Curation of the data, particularly unstructured data, is fundamentally a new skill that businesses have to learn to make LLMs more than just incredibly creative hallucination engines,” Jones says because, “I think [AI is] great technology, but there's never been a technology that's so easy to implement badly.” Taking the time to get data right will pay off in the end. 

Knowledge graphs will be central to human-machine interaction

Knowledge graphs represent a major opportunity for businesses to make the most of AI. These digital representations of human-created knowledge allow AI systems to understand and organize complex information in a way that is useful for decision-making. Muglia sees knowledge graphs as central to AI’s future, enabling businesses to structure unorganized data and drive smarter decisions. 

“A knowledge graph is simply a digital representation of human-created knowledge that can be understood by a machine,” Muglia shares. For example, tables in documents often don't consist of simple rows and columns. They may have sub-columns that need to be translated into a format machines can process. The goal is to create a structure that mirrors how a human would interpret the data.

Muglia suggests, “Knowledge graphs are going to play a very key role, and companies will begin to think more about how to establish their business process and rules in the form of knowledge graphs that can be understood by both people and by machines.” And when the two work together, it could open up insights into parts of the business previously unreached. 

Looking ahead: Predictions for 2025

The conversation concluded with predictions about where AI is headed:

• George Fraser: AI workloads will increasingly resemble business intelligence rather than application development. Platform vendors are creating some great building blocks, and companies are going to find a lot of success in assembling those building blocks together along with a very curated view of their own data in order to create powerful internal tools.
• Bob Muglia: In the next 12 months, knowledge bases will become mainstream, enabling businesses to leverage AI for both internal and customer-facing applications.
• Steve Jones’s prediction: Purpose-specific AI models will outperform general models. He also predicts the rise of a new skill set: professionals who understand both data and its role in operational decision-making.  

Whatever the future does hold, data will be key. Ensuring a strong data foundation for AI will enable better, faster insights that can take organizations to the next level. 

[CTA_MODULE]