Unlocking the full potential of generative AI with data

After looking at the different types of AI in the first part of our series, taking a closer look at generative AI and its special features in the second part, looking at the different models in the third part, and finding out which properties of language models we need to consider when prompting in the fourth part, learning about the practice of prompting in the fifth part, this part deals with the connection between generative AI and data.

We can do a lot with the right prompts. Multilingual websites are suddenly no longer a waste of time or money because AI takes care of the translations. Podcasts can be edited after the fact without speakers having to go back to the microphone. Efficiency gains are immediately possible. But they are not a unique selling point. In the future, almost every company will use AI tools in the same way that almost every company uses digital communication channels today. So what will it take to transfer your own USPs into the future?

Data plays a central role. Every company has its own data that represents a significant part of its expertise. This data can be divided into different categories, such as customer data, market data, competitor data, or internal company data. This distinction is important because each of these categories has a specific function and can be used specifically for the application of AI.

It is often the case that companies have large amounts of data, but it is not always available in a form that can be used immediately. Sometimes there is a lack of awareness of what data is already available. A lot of data sits in archives, is unstructured, or is distributed across multiple systems. But that does not mean it is worthless. If we tap into it, there are numerous ways to use it together with AI to optimize processes or create entirely new products.

Data is the key

So what are some of the products or processes where data is the gold standard for generative AI? And vice versa, by the way.

• Specialized solutions: Combining AI with proprietary data can create custom applications that are different from off-the-shelf solutions. For example, combining customer data with generative AI models can create personalized services or products that are more tailored to the needs of the target audience than generic offerings. Let's take a media company. It could, for example, develop specialized analytics tools to better understand the media consumption of specific audiences and use these results to optimize content. In this case, understanding your own data becomes a critical factor in developing such solutions.
• Work with technology providers: Not every organization has the resources to train its own AI models from scratch. With the right partner ecosystem, solutions can be developed that require such training. This is about being able to assess how and with whom data can be shared and used to implement AI-based applications. For example, a mobility service provider that collects data on the use of its vehicles can work with a technology provider to create predictive maintenance plans. In cases like this, leveraging data and the right technology partner are key to realizing the full potential of your own data.
• Process optimization: Another way to use AI and data to gain a competitive advantage is to optimize internal processes. This requires an understanding of which processes can be improved and how AI can be used to automate them. For example, agencies could use their project and resource management data to identify bottlenecks early and optimize workflow. AI-powered systems could use historical project data to identify typical delays and automatically suggest how to distribute tasks more efficiently. The key is to develop an understanding of what operational data is available and what an ideal process might look like to make operations smoother and more profitable.

The ability to combine data with AI in a meaningful way will be a major competitive differentiator in the future. Knowing how to categorize, use, and refine data is therefore critical-not just for specialists, but for everyone who is responsible for the business.

The next part of our series will focus on how to work with internal data and AI.