your123 - stock.adobe.com
At a time when enterprise IT vendors are tripping over themselves trying to embed the latest AI technologies into their platforms, SAP turned to an old hand: IBM Watson.
SAP has signed a partnership with IBM that embeds IBM Watson into SAP Start, a search interface for SAP cloud applications and SAP S/4HANA Cloud.
The new IBM Watson integration is intended to boost users' productivity with both natural language capabilities and predictive insights.
"This milestone collaboration with IBM aims to provide SAP customers a better user experience, faster decision-making and greater insights to help transform their business processes," SAP CEO Christian Klein said in a press release.
The new digital assistant capabilities in SAP Start will be extended across SAP applications to help users answer questions as well as automate or speed up common tasks, according to the company. IBM Watson is currently available in SAP Concur's TripIt mobile app to assist with travel-related decisions.
It's likely more AI-related news will drop for SAP. During a first quarter 2023 earnings call in April, Klein teased that ChatGPT could make an appearance at the SAP Sapphire conference.
Jumping on the AI bandwagon
SAP has been trying to take advantage of IBM Watson capabilities for a while, as shown in the SAP Concur integration. It now looks like it is trying to expand these capabilities to other industry use cases, said R. "Ray" Wang, founder and analyst at Constellation Research.
"This is SAP's generative AI moment where they're taking advantage of AI and answering questions around whether SAP has AI-enabled solutions," he said. "Why not go with one of your biggest partners who has been doing this for a while and use that to start the conversation around AI and what they can deliver inside the SAP system?"
Joshua GreenbaumPrincipal, Enterprise Applications Consulting
However, the partnership looks more like it's meant to assure investors that SAP is in on the AI bandwagon, said Joshua Greenbaum, principal at Enterprise Applications Consulting.
IBM Watson is a "failed technology" that enterprise customers have shown little interest in over the years, he said.
AI in general doesn't look like a high priority for SAP customers, Greenbaum said, noting a recent survey from Americas' SAP Users' Group members, which showed AI and machine learning were eleventh in a list of important technologies.
Nevertheless, it appears that SAP intends to prove it has a commitment to integrating AI, he said.
"They are certainly trying to lay up a series of AI ducks in a row with Sapphire coming up. But this is a message to let Wall Street know that they're doing this stuff," Greenbaum said. "It's blockchain and crypto all over again. It's not going to change the revenue stream or customer success; it's about making SAP look good to investors."
IBM Watson needs to prove its case
IBM Watson was built as a natural language processing and knowledge modeling system that debuted in 2011 on the quiz show Jeopardy!, where it defeated two former champions. IBM subsequently launched it commercially as a product in healthcare, but Watson was never widely adopted.
IBM overestimated Watson's capabilities to provide answers healthcare researchers needed in areas like cancer research, said Jon Reed, co-founder of Diginomica, an enterprise industry-focused analysis site.
IBM Watson gets a bit of a bad rap as IBM has worked on developing useful capabilities for it, including applications for weather forecasting, tax preparation and advertising.
"They took an ambitious tack with Watson but then stepped back and started to use it in a more practical way," he said.
SAP will need to work on finding ways to make these capabilities work for its customers.
"SAP is going to work on its own AI development," Reed said. "But even if IBM has had these high-profile difficulties around Watson, they've still been working on it for a longtime, which gives them a good maturity level around some of the tools."
For example, IBM Watson has been used for HR applications in providing a self-service interface that lets users streamline and simplify tasks such as onboarding, he said.
"Some people will find it useful to have an interactive front end on top of applications, but it's not certain that everyone's going to like it, so it will depend somewhat on people's preferences and how good their enterprise search capabilities are," Reed said. "That may be where SAP sees opportunities to make the SAP Start application environment easier to navigate."
Generative AI will undoubtedly be a significant topic at Sapphire. However, SAP's challenge will be to distinguish its offerings from all the other AI rollouts enterprise vendors are making, he said.
"On the one hand, it's reassuring for customers to know that their trusted partners are investing in AI," Reed said. "But on the other hand, it creates a big evaluation problem to figure out which vendors are furthest along, which have the most useful applications and what the customer's ROI will be."
Jim O'Donnell is a senior news writer who covers ERP and other enterprise applications for TechTarget Editorial.