SAP Databricks is integrated directly into SAP Business Data Cloud (BDC) as an embedded component.
SAP Databricks provides a streamlined way to access, analyze, and leverage your SAP data using Databricks' advanced AI and analytics capabilities. SAP data that has been replicated from SAP applications to SAP Business Data Cloud is shared using zero-copy with SAP Databricks through data products. This eliminates the need for separate, complex data extraction, transformation, and loading (ETL) processes from SAP applications to SAP Databricks, This saves significant time, reduces complexity and uses fewer resources.
Key Features of SAP Databricks
SAP Databricks includes a wide range of tools to support the data scientist. Let's highlight two of them.
Explanatory Data Analysis
Consider a new dataset as the pieces of a puzzle you need to solve. Before you can solve this puzzle, such as building a predictive model, you must first look at the big picture, identify the different shapes and colors so you can group the pieces that appear to be related. This initial investigation is called Exploratory Data Analysis (EDA).
EDA is the crucial first step in any data science project. It's an approach where we use visualizations and statistical summaries to understand a dataset's main characteristics, uncover hidden patterns, spot anomalies, and test our initial assumptions. It’s less about formal hypothesis testing and more about asking: "What can this data tell me?".
The goal of EDA is to develop a thorough understanding of your data so that you are able to choose the best algorithm, settings and parameters to suit your data.
For example, you could use EDA to identify the distribution of customers across countries. If you discover there are many countries in the data set with some countries having very few customers, you might then decide to group the countries by continent and apply different modeling settings per continent.
Or you could use EDA to investigate the sales revenues over a specific time period to determine if the sales for individual months should be removed from the analysis. You might want to remove data because an exceptional event affected the sales revenue and this data should not be included in the data modeling.
Using SAP Databricks, you can perform EDA more efficiently by collaborating on data science projects using interactive notebooks. Notebooks support multiple programming languages such as SQL and Python. These notebooks facilitate data exploration, visualization, feature engineering, and model development.
Automated Forecasting
A key feature of SAP Databricks is the automated forecasting. Automated forecasting is a powerful approach that uses artificial intelligence (AI) and machine learning (ML) to predict future outcomes based on historical data. At its heart, this technology feeds vast amounts of historical information into sophisticated algorithms that work automatically to identify hidden patterns, trends, and relationships that a human analyst might miss. By doing so, it streamlines the entire forecasting process, allowing organizations to generate highly accurate predictions with incredible speed and efficiency, moving beyond the limitations and time constraints of traditional methods.
With SAP Databricks, you can build accurate time-series forecasts using advanced algorithms and automated hyperparameter tuning.
Key Features of SAP Databricks
Discover the key features of SAP Databricks in the following video.
Use Cases for SAP Databricks
Leveraging the SAP data products of SAP Business Data Cloud with the addition of 3rd party data, SAP Databricks supports many business use cases across all lines of business and industries.
Here are some examples:
- Supply Chain Risk Modeling - Anticipate disruptions by analyzing supplier performance trends and geopolitical factors, transportation delays using weather/port congestion data, and alternative sourcing scenarios to maintain production continuity.
- Dynamic Pricing - Use predictive analytics to adjust prices in real time based on market trends, competitor actions, and customer demand.
- Cash Flow and Liquidity Forecasting - Use historical payables, receivables, and payment behavior to forecast cash positions accurately, identify potential liquidity issues, and optimize working capital management.
- Demand Forecasting - Use AI and analytics to accurately predict future demand, enabling proactive production and supply chain planning.
Summary
- SAP Databricks seamlessly integrates Databricks' AI and analytics capabilities with your SAP landscape.
- It provides zero-copy access to SAP Business Data Cloud data products via delta sharing, ensuring data consistency.
- The serverless architecture simplifies provisioning and reduces infrastructure management overhead.
- SAP Databricks includes tools to support the data scientist including exploratory data analysis and automated forecasting.