Data Attribute Recommendation
Perform classification and regression tasks with this machine learning-based service available on SAP Business Technology Platform.
SAP is introducing a series of generative AI capabilities and advancements aimed at empowering developers of all skill levels to supercharge their businesses in the age of AI. It includes SAP Build Code, a new vector capability in SAP HANA Cloud, a one-stop-shop for SAP BTP developers called AI Foundation, and much more!
With the introduction of Transfer Learning, you can now update a trained classification model by simply uploading new data and triggering a training run to incorporate the changes. With ever-increasing data volumes, this not only saves time when updating a machine learning model but also provides more flexibility in the model’s lifecycle.
SAP BTP acts as the enabler for SAP and partner applications to leverage generative AI capabilities in an SAP context. Generative AI capabilities in SAP BTP will improve the developer experience, how integrations are done, the process automation space, analytics and holistic planning activities, as well as simplify administrative tasks.
It uses free text, numbers and categories as input to classify entities such as products, stores and users into multiple classes and also to predict the value of missing numerical attributes in your data records.
With Data Attribute Recommendation, companies can:
- Use classification and regression capabilities to speed up data management processes
- Reduce errors and manual efforts for completing business processes
- Increase data consistency and accuracy
Be flexible in uploading and storing data in separate datasets.
Model training metrics
Expose model training metrics after a successful training run.
Pick one of multiple machine learning models that best suits your needs.
AutoML model template
Choose the best fitting machine learning pipeline for single label classification tasks.
Deploy multiple models at the same time while having the flexibility to configure the inference response.
Deploy Data Attribute Recommendation Capabilities
Users can easily deploy Data Attribute Recommendation in several business scenarios and extend their own processes.
Perform a regression on the data to predict a numerical label
Upload a dataset to your Data Attribute Recommendation service instance to afterwards be able to train your machine learning model using the regression model template.
Follow the tutorial
Sales order completion
A business blueprint with dedicated preprocessing for autocompletion of sales orders. The underlying machine learning model can predict custom, defined incompletion log fields, provide custom metrics, and provide a training pipeline enabling the user to define which data should be used for training, testing, and validation to overcome issues with imbalanced data sets.
Start the tutorial
Available Integrations in SAP's Portfolio
Data Attribute Recommendation is natively integrated into several SAP solutions.
Intelligent recommendation of material groups for centrally managed demands in SAP S/4HANA Cloud
Central purchasers can proactively receive intelligent proposals for incorrectly assigned material groups within their centrally managed purchase requisitions.
Watch the demo
Read the blog post
Smart account determination in SAP Central Invoice Management
Automatically enrich G/L Account, Cost Center and WBS Element fields in draft invoices using the data learned during the training, and increase accuracy of financial reporting and reduce manual work.
SAP Intelligent Product Recommendation
Cloud application that uses machine learning and rules to streamline the product selection and configuration process for complex configurable products. It recommends products and default configurations based upon the customer's specific needs. You can embed SAP Intelligent Product Recommendation in configure, price, quote (CPQ), or e-commerce business processes. It also supports single-level configurable products as well as integration to SAP S/4HANA as important sources of sales transactional data.
By utilizing historical data and machine learning, the system is now able to provide recommendations for completing empty fields within an unfinished sales order, increasing sales productivity while reducing data management and operational costs.
Watch the demo