Technology Blogs by SAP
Learn how to extend and personalize SAP applications. Follow the SAP technology blog for insights into SAP BTP, ABAP, SAP Analytics Cloud, SAP HANA, and more.
cancel
Showing results for 
Search instead for 
Did you mean: 
susenpoppe
Product and Topic Expert
Product and Topic Expert
665

Key Take Aways


Multi-model databases simplify data management by handling various data types in one system, reducing complexity, and requiring a leaner skillset. They enable unified data integration, advanced analytics, personalized recommendations, real-time decision-making, and IoT applications.

SAP HANA Cloud offers specific engines for different data models, such as a Graph Engine for relationships, a JSON Document Store for semi-structured data, and a Spatial Engine for geospatial analysis. These empower users to build Intelligent Data Applications without specialized solutions, promoting holistic data-driven decision-making.

Introduction


Data is omnipresent. In 2025 we will create, capture, copy and consume 181 zettabytes of data.[1] One Zettabyte is around 1 billion Terabytes.[2] This data explosion makes it crucial for businesses to tackle both data volume and data variety in the most efficient way.

This blogpost dives deeper into the use of multi-model technologies within Intelligent Data Applications and the role SAP HANA Cloud’s capabilities play in making the vast amount of data usable. Also, please be sure to watch the interview with the SAP HANA Cloud Product Specialist shabana31 During this fifteen-minute video, Shabana shares valuable insights into the world of multi-model data processing, potential use cases, and real-world customer examples.


Interview with Shabana Samsudheen on multi-model powering Intelligent Data Applications.



Multi-model


Multi-model refers to a database’s capabilities to support different types of data in a single system. Common examples include relational data, objects, documents, graphs, and more. The exact combination depends on the vendor. As multi-model databases do not rely on storing data in row-based tables solely, they can handle not only structured but also semi-structured and unstructured data.[3]

Multi-model, not multiple models


Today’s Software as a Service (SaaS) applications require support of different data models and workload types. Instead of using multiple highly specialized database technologies that require an infrastructure layer, multi-model databases offer a simpler solution. They reduce the complexity down to a single database technology. [4] These unified database platforms are capable of meeting different application requirements and eliminate the necessity for niche database systems.[5] Relying on a single solution boosts cost efficiency while development and maintenance become much more feasible. Of course, infrastructure costs reduces as the infrastructure layer becomes significantly less complex.

Especially in the IT realm, the shortage of skilled workers makes hiring difficult. Maintaining a multi-model database requires a much leaner skillset, as opposed to the skillset required to take care of various special purpose solutions.

Additionally, if one data model becomes redundant, the migration within a multi-model database is less of a hassle. Things like transfer costs, downtime, security in transit, and data loss are not a problem with intra-database migrations.

Businesses must deal with various data formats coming in from different sources. Using a multi-model database, that supports different data models eases this task. It allows users to look at and analyze data coming from different sources simultaneously. This also means less ETL (extract, transform, load) operations are needed as the different data models are natively integrated on one platform.

Common Multi-model Use Cases


There is a broad variety of use cases showcasing the usage of multi-model capabilities. Successful projects build Intelligent Data Applications that leverage multiple data sources while delivering analytics, insights, and automation across various industries and business scenarios.

Unified Data Integration

Customers leverage multi-model capabilities to first integrate data from different sources and models, to then unify them into one view. Consolidating not only structured but also unstructured, graph, and geospatial data, they create a comprehensive data foundation for their Intelligent Data Applications.

Intelligent Analytics

Performing advanced analytics to gain actionable insights is crucial for businesses competing in today’s fast paced economy. A multi-model database allows them to leverage…

… a relational model to perform complex queries and aggregations.

… a document store to analyze unstructured data.

… a graph model to identify relationships & patterns.

… a spatial model to do geospatial analysis of the data.

By uniting the individual strengths of the different data models, users can uncover hidden correlations, make data-driven decisions, and drive innovation.

Personalized Recommendations

Multi-model databases allow users to use graph models and machine learning algorithms to deliver personalized recommendations. (Stay tuned to part 3 of this series to learn more about machine learning). By analyzing user behavior, preferences, and connections, application builders can provide suggestions for products, contents, or services, as well as enhance the user experience and drive engagement.

Real-time decision-making

Combined with in-memory computing, multi-model databases enable Intelligent Data Applications to support real-time decision making. Processing and analyzing data immediately allow real-time anomaly detection, the triggering of automated actions, and responding to changing conditions on the fly. Real-time decision making helps to improve operational efficiency and customer satisfaction.

Internet of Things (IoT) Applications

Especially regarding IoT, multi-model plays a vital role. Customers can use a document model to store and analyze sensor data. Graph capabilities help to identify relationships and anomalies. Geospatial analysis allows the user to gain insights from location based IoT data. Finally, these applications can even feature operation optimization, pattern detection, and predictive maintenance.

Graph, JSON Document Store, Spatial – SAP HANA Cloud’s multi-model engines


SAP HANA Cloud offers three specific engines catering to different data model needs.

Graph Engine

The graph engine analyses the many relationships existing between data entities e.g., raw good suppliers for a manufacturing company. These relationships are essentially networks allowing for easy traversal. As an example, one could perform an impact analysis on the supply chain.

The standard interface to use SAP HANA Cloud’s Graph engine is SQL which simplifies the use within applications. Additionally, SAP HANA Cloud offers graph script language for writing in-database graph script procedures. Instead of tables and columns that are used in SQL scripts, graph script uses edges, vertices, paths, and sub graphs. Several popular algorithms, like finding the shortest path or page rank are already built in SAP HANA Cloud.

JSON Document Store

To handle semi-structured data, SAP HANA Cloud supports the JSON data model. Therefore, users can store and retrieve data e.g., from documents, IoT sensors, or social media. They are also able to perform CRUD (create, read, update, delete) operations on this type of data. Being fully schema-flexible and supporting nested objects and arrays, data can be stored without defining a schema upfront.

Spatial Engine

The spatial engine analyzes data using points and other geometric objects. This means users can store geospatial data like GPS coordinates or geospatial features as well perform spatial analysis, proximity searches or location-based queries.

The standard interface to use the spatial functions is SQL. Hence, collaboration with any other database application becomes simple. Being natively built into the database, the spatial functions can profit from SAP HANA Cloud’s famous in-memory technology.

Offering support for all these different data types, SAP HANA Cloud gives the freedom to pick and choose the right engine that fits the use case best. Specialized solutions are no longer needed. Data can be combined and integrated across all types, enabling advanced analytics, comprehensive insights, and holistic data-driven decision-making. Thus, users are empowered to build truly Intelligent Data Applications.

The below video demonstrates how these capabilities work together in the context of a beauty retail store's target marketing strategy. Spatial analysis is used to identify potential locations for new stores and analyze market gaps. The use of graph analysis allows one to discover the most popular rated products using the PageRank algorithm. Importing and working with JSON data is seamless and gives the opportunity to look at the possibility to craft personalized marketing messages for target customers. Hence, SAP HANA Cloud enables users to derive valuable insights and drive effective target marketing efforts without complex data transformations – offering a streamlined and efficient solution.


Demo multi-model in SAP HANA Cloud.



Real-world Examples for Intelligent Data Applications using multi-model


A prominent use case is risk management. One customer built an application to mitigate and forecast risks. Handling financial and non-financial data as well as enabling automated risk simulation are key parts of the application. The key technology components were SAP Business Application studio for application development, SAP HANA Cloud for storing the various data models coming from the financial and non-financial risk data, and SAP Analytics Cloud for advanced dashboarding and reporting capabilities.

Another example of Intelligent Data Applications using SAP HANA Cloud’s multi-model capabilities can be found in the utilities sector where in-depth spatial insights are mandatory. Customers benefit from the multi-model capabilities that combine operational data with field assets to deliver actionable business intelligence.

Additionally, the use of advanced multi-model analysis can be seen in sports and entertainment where applications have been built to derive gaming strategies from real-time data insights. Once again, this example showcases how SAP HANA Cloud customers benefit from a solution that utilizes both in-memory storage and multi-model capabilities.

Summary


Multi-model databases, like SAP HANA Cloud, play an important part in enabling businesses users to succeed in today’s fast paced economy. Customers are benefiting from a database that provides a unified data model, advanced data analysis, and in–memory speed by building intelligence into their applications. Thus, Intelligent Data Applications using SAP HANA Cloud can deliver solutions that span every industry and all lines of business.

Great Resources to start today



  • Experience SAP HANA Cloud today

  • Learn how to import data and schema with SAP HANA database explorer

  • Understand how to create graph workspace on SAP HANA Cloud

  • Find more information about the graph workspace in SAP HANA Cloud

  • Gain a better understanding of SAP HANA Cloud’s multi-model engines

  • Check out daniel.dukes blog post demonstrating a unique use case for an intelligent data application combining HANA Cloud with solutions from Esri and HERE


 

Sources


[1] https://www.statista.com/statistics/871513/worldwide-data-created/

[2] https://www.computerweekly.com/de/definition/Zettabyte#:~:text=Zettabyte%20(ZB%2C%20ZByte)&text=Die%...

[3] https://www.techtarget.com/searchdatamanagement/definition/multimodel-database#:~:text=A%20multimode...

[4] https://www.singlestore.com/blog/what-is-a-multi-model-database/

[5] https://www.techtarget.com/searchdatamanagement/definition/multimodel-database#:~:text=A%20multimode...