SAP's AI Future

In fact, SAP has been employing AI for years.

To learn more, UKISUG hosted a podcast with Timo Elliott and Vivien Boche to discuss how SAP is currently using AI, what the future uses cases might be, and the pitfalls we all need to look out for. The current state of play and where we are headed

SAP products using AI that you might not know about include Concur, which uses machine learning to scan receipts, allowing the application to “read” and input data. Another example is SAP’s predictive analytics libraries in HANA cloud, which use AI to give users better insights on their data. SAP also has plans in the pipeline for more use cases, particularly to enhance analytics. For example, using natural language AI so users can ask questions of their data in everyday language, with the AI choosing how to present the data so its easily understood. The result is data that is better presented to users, making analytics tools easier to use.

The ethics conundrum

Another topic of discussion concerned the importance of ensuring AI is applied ethically and responsibly. This is something many of you are probably concerned about. Timo highlighted that SAP’s algorithms had been specifically checked to ensure they didn’t hold bias towards race, gender or even technology. AI “hallucination” is another challenge to be overcome. This is where an AI can give false information without knowing it’s false – something Vivien warned can happen readily when using generative AI.

Concerns around hallucination or bias don’t mean we should shun this new technology. Used safely, there are some fascinating applications for generative AI. For instance, CAPGPT – Cloud Application Programming Model GPT – allows users to create cloud applications by simply prompting the AI in normal language. It can also be used to generate test data to check if an application will function as planned – a task that can be hugely time consuming when undertaken manually.

AI is clearly here to stay, and will be an area SAP continues to invest in. We look forward to seeing how the use of AI develops over the coming years, and we have no doubt we’ll be discussing it again soon. Listen to the full podcast here.