AI is becoming a topic of discussion in strategic meetings worldwide, but actual implementation hasn't kept pace with expectations. According to Irfan Khan, President and Chief Product Officer of SAP Data & Analytics, many businesses are now 'more worried than active' when it comes to AI.
The root cause lies not in technology, but in the data infrastructure. Most businesses are still failing to leverage their data effectively because critical information is scattered across multiple different systems.
By 2026, the market will be clearly divided into two groups: businesses that are already grounded for AI and those left behind. A SAP survey revealed that only 33% of leaders truly trust their data. This trust gap leads to a gap in deployment — and companies ahead are up to three times faster in getting ready for AI.
Here are four crucial steps that businesses need to take immediately.
1. Embrace hybrid cloud, but don't let your data be fragmented.
Despite years of migrating to the cloud, most businesses are still not fully 'cloud-native'. Legacy systems and legal constraints make a hybrid environment (combining cloud and on-premise) almost inevitable.
However, AI doesn't require a complete overhaul of the infrastructure. The key is that businesses can access data consistently and securely, regardless of where the data is located.
That's why SAP is promoting the 'data fabric' model — a connecting layer between data sources and AI applications, providing the necessary business context for AI to work effectively.
For example, Yamaha—a company with an extremely diverse product ecosystem, from musical instruments to motorcycles—can connect product, supply chain, financial, and customer data thanks to the SAP Business Data Cloud platform, thereby accelerating decision-making in the AI era.
2. Building a data system 'designed to serve AI'
Traditional databases are not designed for AI. Therefore, businesses need a new data platform that can handle multiple data types simultaneously, from vector and knowledge graphs to real-time data.
SAP has chosen a multimodal approach, allowing AI to query data directly without copying or duplicating.
A prime example is the agricultural company Martinez and Valdivieso in Chile. When implementing SAP HANA Cloud, they equipped their sales team with AI tools that could generate real-time quotes based on data from multiple sources.
As a result, the sales process becomes faster, more accurate, and provides a better customer experience.
3. Focus on where AI creates real value.
A common mistake is trying to implement AI everywhere at once, leading to a dispersion of resources. According to Irfan Khan, businesses should start with specific use cases—especially those with high risk and uncertainty.
The supply chain is a prime example. Forecasting supply and demand is always a difficult task, and this is where AI can make a significant impact.
Instead of reacting after an event occurs, AI enables businesses to anticipate and act early. This approach can also be applied to finance, operations, human resources, and many other areas.
4. Restructuring the team and operating model.
AI is not only changing technology, but also changing the way people work.
In the short term, businesses will operate on a 'human + AI' model. AI handles repetitive tasks such as testing or quality control, while humans focus on innovation and decision-making.
In the long term, traditional operating models like waterfall or agile will also have to change to adapt to the speed and flexibility of AI.
This requires businesses to eliminate outdated data systems and switch to a more flexible platform, often cloud or hybrid.
The most important message isn't about 'whether or not to use AI', but rather 'are businesses ready for AI'.
If you're still struggling to find your 'business case,' chances are your business is already behind the market. In this new era, rapid experimentation, rapid failure, and rapid learning will become core competitive advantages.