How far will the flywheel of data and AI carry us?

One of the richest sources of business information has been hiding in plain sight. It’s a veritable gold mine of business value and insights, yet it has always lain just out of reach. Until now.?

I’m talking, of course, about unstructured data. That is, all the customer conversations, social media chats, call logs, video recordings, whiteboard brainstorming sessions, case notes, and more that enterprises might collect and store, but haven’t considered as part of the enterprise data landscape.?

Before, it was too hard to combine all these data signals with other enterprise data to derive insights. Now, thanks to generative AI, it’s easy. The latest multimodal models can understand, operate across, and combine different types of information—including text, code, audio, image, and video.?

More accessible than ever before, generative AI gives everyone the ability to activate the largest, unused data set in existence today. No wonder, then, that a key prediction in Google Cloud’s 2024 Data and AI Trends Report was that generative AI will speed the delivery of business insights across organizations.

I’m excited about what lies ahead. And it makes me think, what can we all do to prepare?

The flywheel of data and AI

To tap into the incredible opportunities that generative AI presents, organizations need to rethink the relationship between data and AI.?

In the past, a great divide existed between data and AI, both technically and role-wise. This was largely driven by system limitations. Now, though, we’re starting to see what’s possible when these two domains blur and generative AI is put to work.

Generative AI agents can now effectively activate data to extract insights, make decisions, and automate processes. All that’s needed is for the AI and the data to be in close proximity —?which means two things. First, you need to make your data easily accessible to your AI models. Then, you also need to make the AI tools and services readily available to data practitioners.?

When you get this right, then the innovation potential is huge. And it’s self-sustaining, just like a flywheel.

For example, think of a warehouse. Instead of just running inventory predictions based on historical data, you can now use an AI model to optimize your restocking and shelving strategy. The flywheel of data and AI opens up whole new scenarios and insights.

And we’re already seeing the magic at work in companies as diverse as a travel company and a financial technology platform . These companies are showing what’s possible when you leverage new data signals — all based on the ability to analyze and understand unstructured data in combination with business data.??

Accuracy is critical

For generative AI to take off in scenarios like those mentioned above, both internal users and end customers need absolute trust in the output. To chat naturally with a generative AI agent, you need to know that the information it shares can be relied upon.?

Just as people around the world trust Google.com as a source of truth for online search, enterprise generative AI apps need to earn such trust. To gain it, these apps need to be grounded in specific business context.?

This is where operational databases and capabilities like Retrieval Augmented Generation (RAG), grounding services and vector embeddings come in. Without getting too technical, these techniques and tools help ensure that enterprise generative AI apps only use accurate, up-to-date, relevant data—which in turn mitigates the risk of hallucinatory or incorrect responses.??

It’s just the beginning

The volume of data in an organization will never be as small as it is today. It’s only going to grow.?

And, to innovate with this data without prohibitive costs or complexity, you need a data platform that’s built for the job. From my perspective, there is no more suitable system for the AI revolution than BigQuery . Built for large-scale data processing and completely serverless, it is a remarkable, one-of-a-kind system that can work seamlessly with structured and unstructured data.?

Most importantly, it’s affordable. TechTarget’s Enterprise Strategy Group (ESG) found that BigQuery customers save up to 54% in TCO compared to other cloud data platforms—which will really help as you start doing more, and more, and more with data.

Thanks to generative AI, the time has come for every enterprise to innovate. With the trends in the 2024 Data and AI Trends Report already coming true, I can’t wait to see what will come next.