Though many organizations are interested in the idea of bringing automated intelligence to their work, few have figured out a practical way to do this yet. There’s no doubt that machine learning can augment the power of the human workforce and help us understand so much more than we can today with our current analytics. The hard part now is knowing what we do not know. But machines, as they get smarter, may be able to tell us.
The Power of Human + Machine Learning
To get the most out of a predictive model, you need to constantly retrain each model and manage its end-to-end life cycle. SAP is aiming to do this based on rules and predefined patterns that machine learning can identify and apply so the system can constantly retrain itself.
For example, if the finance department mistakenly receives two invoices for the same product, the system can use machine learning to not only catch this error, but also learn how to find it again. No need to hire a data scientist or mathematician to reprogram the algorithm—the system learns automatically for you. This is how we can all train the machine brains to go to work for us, massively expanding the speed and scope at which we can currently accomplish our daily tasks.
Where humans can bring in their knowledge and control is in building the predictive models in the first place. You can create multiple models, each based on a specific hypothesis, and then test them in your systems in real time. As you compare them with actual data, you can then decide which of the models is the best one to use.
Back to Machine Brain School
When it comes to machine learning, it’s a constant case of back to school if we want these new tools to remain effective. So where will all of this training material in the form of data come from?
SAP has a breadth of internal and partner use cases to draw on to feed the machine brains. SAP explains that it is now working with major enterprise partners, including Accenture, to identify the use cases that these machine learning features will need to draw their intelligence from. This specific work is focused on information drawn from SAP S/4HANA, SAP S/4HANA Cloud Edition, SAP Hybris, SAP Ariba, and more products to come.
How to Learn Without Perfect, Clean Data
When building models without the help of machine learning tools like SAP’s, you need perfect, clean data from your enterprise to run them. If you really want to forecast something, you need to start with accurate and stable data. Otherwise you’ll find the truth in the old “garbage in, garbage out” adage. It’s not realistic for most companies to find this type of data in their systems without having to do a lot of preparation to clean it.
That’s where machine learning powered by real data from real use cases comes in. The system will use rules and predefined patterns to correct the data as it encounters errors, inaccuracies, and omissions. Meanwhile, it takes the best of the real-world data and uses that as a reference point as it learns.
Democratizing Machine Learning Across Applications
SAP is engineering this more-accessible version of machine learning intelligence for predictive analytics functions, available through SAP Predictive Analytics App Edition. Its mission is clear: SAP wants to democratize machine learning across applications.
SAP Hybris already has machine learning intelligence for predictive analytics built in. A previous third-party recommendation engine has now been replaced with an SAP Hybris engine that is specifically tuned to commerce and marketing to help with actions such as targeting and segmentation using the predictive model. If users wish to extend its power, then they can do so through SAP Predictive Analytics App Edition.
As SAP continues to augment SAP Leonardo with its growing number of industry accelerators, it will be interesting to see how task-specific machine learning enhancements get smarter in their subject-matter expertise. These capabilities could certainly find their way into an SAP Leonardo project, and the SAP Predictive Analytics App Edition could fit right within a company’s associated blueprint.
Building the Machine Brains Right from the Start
If artificial intelligence (AI) lives up to its promise of real-world business impact, then it’s clear that we need to build the machine brains right from the start and make their intelligence as widely accessible as possible throughout an organization. If we do this well now, the future will be smarter, in a machine learning kind of way.
Sign up for our SAP Leonardo webcast series to hear more about how you can benefit from the latest advancements in machine learning and artificial intelligence.