SAP will advance their existing machine learning capabilities of cloud for robotics process automation.
SAP today revealed it will be taking advantage of machine learning algorithms to embed robotic process automation (RPA) functionality.
Announced at the SAP Tech-Ed Europe 2018 conference, the RPA capabilities will be made available as part of the framework of advanced technologies that SAP makes available as part of the SAP Leonardo initiative. SAP envisions those RPA capabilities will be employed to automate a range of business processes using bots that leverage conversational artificial intelligence (AI) capabilities that SAP has developed.
RPA is already widely employed to automate a range of responses to events that typically involve the processing of documents. By taking advantage of machine learning algorithms SAP beginning in early 2019 expects to be advance the RPA by, for example, making it simpler to completely automate the processing of invoices, says Markus Noga, senior vice president for machine learning at SAP.
Those machine learning algorithms would also be able to more readily identify potential instances of fraud that continue to plague vertical industries such as the healthcare sector.
As part of this effort, SAP is also making it simpler for organizations to create their own AI models based on machine learning algorithms. Working in collaboration with Google, SAP announced today five additional ML services, including a scene text recognition service that extracts text from arbitrary image contents and scenes, customizable image segmentation using pixel-wise classification to enable the detection and recognition of objects in pictures and their shapes, and a customizable text feature extraction service to detect fine-grained patterns in text and enable matching of similar documents.
The two other services are based on a cloud text-to-speech machine learning library from Google that converts written text to speech to power conversational applications and cloud speech-to-text service that converts speech to written text using Google’s DeepMind WaveNet technology to power conversational applications. Those services will be able on Google Cloud Platform (GCP) next month.
Finally, SAP today also announced a software development kit and associated application programming interfaces (APIs) for SAP Analytics Cloud that takes advantage of machine learning algorithms embedded in the cloud service to create more advanced analytics applications and dashboards.
SAP is pursuing a dual strategy when it comes to machine learning. On the one hand, SAP is embedding machine learning algorithms within its entire application portfolio, while at the same time making machine learning algorithms available as a cloud service that developers can readily invoke within custom applications.
“You can think of it as data science in a box,” says Noga.
SAP is betting that investments in machine learning algorithms, RPA, and advanced analytics will drive a wave of upgrades across its entire application portfolio to create what the company describes as an intelligent enterprise. SAP hopes most of the usage of those services will come via packaged applications running on a cloud it manages on behalf of customers. But SAP is also still committed to making those capabilities available in on-premises environments. It’s just every new capability SAP now develops will be made available first on an SAP cloud service.
It’s not clear how long it will take for the intelligent enterprise to fully manifest itself. But what is clear is the time between when an event occurs and a process gets kicked off in response will soon be measured in seconds.