SAP today at its virtual TechEd 2024 conference added a bevy of artificial intelligence (AI) capabilities, including generative AI agents that have been trained to automate a wide range of complex tasks.
The latest iteration of the Joule co-pilot from SAP adds an ability to leverage generative AI to automate much more complex tasks, says Mohammed Alam, a member of the Executive Board of SAP SE responsible for product engineering.
In effect, rather than simply being able to interact with SAP applications via a natural language interface, organizations starting this quarter via Joule will be able to invoke multiple AI agents. “Those agents will be able to perform multiple actions,” he says.
The first two agents provided by SAP will enable organizations to employ autonomous AI agents to analyze and resolve a range of dispute resolution scenarios, including incorrect and missing invoices, unapplied credits and denied or duplicate payments, and a financial accounting agent that streamline financial processes such a bill payment.
Ultimately, Joule agents will, over the course of the next several months, be embedded into every major SAP platform, including by the end of the first quarter of 2025 Joule agents will translate natural language inquiries into insights using real-time data and a set of agents that will be embedding into the SAP RISE managed service.
As part of that effort, SAP also revealed it has SAP announces expanded its AI alliances with Anthropic, IBM, Meta and Mistral in addition to expanding services partnerships with HCLTech and PwC.
The overall goal is to dramatically increase productivity involving a wide range of tasks, including software development.
For example, application developers via the SAP Build platform for building custom applications will in 2025 be able to use Joule to create applications written in ABAP, the programming language most custom SAP applications were developed using. Joule will also be able to generate explanations for legacy code, making it easier to modernize codebases. Starting this quarter, capabilities such as code explanation and documentation search in SAP Build Code will also reduce development time for Java and JavaScript developers in addition to being able to use Joule to automate workflows in the SAP Build Process Automation platform.
In addition, Joule before the end of this year, will be integrated into SAP Build Work Zone, a low-code tool for creating applications. In the first quarter of 2025, application developers will also be able to extend Joule studio, a tool for extending Joule, to both SAP and non-SAP applications.
SAP in the first half of 2025 will also make available a SAP HANA Cloud knowledge graph engine based on its core database platform to make it simpler to manage the data sources required to build and extend large language models (LLMs), noted Alam. The SAP knowledge graph provides a semantic layer that connects ABAP tables, views, applications and application programming interfaces (APIs).
Finally, SAP is making improvements to the generative AI hub capability in SAP AI Core and SAP AI Launchpad by the end of Q4 2024 to better enable developers to build, customize and deploy complex AI applications using a wider range of foundational models.
Just about every organization is debating now to what degree can they simply enjoy the benefits of AI as a feature that is embedded within an application, versus having to build and deploy their own AI models. SAP is clearly looking to provide both options to organizations using a set of applications that can be customized and extended in ways an organization might best see fit.
