The Challenge
Enterprises today work with massive volumes of data spread across tables, views, JSON collections, graphs, and text. Yet most users don’t know where that data lives or how to access it efficiently. Building a bridge between business questions and the underlying technical data model often requires deep schema knowledge and SQL expertise.
At the same time, users expect to “talk to their data”, to search, ask, and get answers without navigating complex structures. But creating such experiences has often been challenging. Data is often siloed, metadata is hard to navigate, and the context needed to connect insights is missing.
The Innovation: Discovery Agent and Data Agent in SAP HANA Cloud (planned QRC1 2026)
At SAP TechEd 2025, SAP HANA Cloud takes a major step forward by becoming agentic and introducing a suite of agents designed to simplify how people discover and interact with data.
In this blog, the focus will be on the Discovery Agent and the Data Agent.
Together, they make it easier for agent users to find the right data objects, understand relationships, and generate the right queries, all without needing deep technical or SQL knowledge.
SAP HANA Cloud Suite of Agents
Discovery Agent: Exploring data context
The Discovery Agent helps users find and understand the data available in SAP HANA Cloud. It explores the HANA catalog using a Knowledge Graph that exposes how objects relate to one another, making it easy to locate relevant datasets without manually browsing schemas. When users describe what they need, like for example, project details or customer transactions, the Discovery Agent can identify the right data sources based on context, relationships, and annotations. Under the hood, the agent generates SPARQL queries on the RDF graph of HANA objects, revealing how data structures connect.
Data Agent: Turning intent into executable queries
Once the relevant data objects are found, the Data Agent takes over.
It translates natural language intent into executable queries, such as SQL. Users can simply describe what they want to retrieve, for instance, “show my current projects with partner involvement” and the Data Agent generates and executes the corresponding query.
By combining discovery and query generation, these agents make data exploration seamless: the Discovery Agent provides context, while the Data Agent delivers answers.
Powered by the SAP HANA Cloud Custom Database Objects Knowledge Graph (planned QRC1 2026)
At the core of these agents lies the Custom Database Objects Knowledge Graph: an RDF graph automatically built from system catalog that captures tables, columns, relationships, and annotations. It provides the semantic layer needed for exploration, AI grounding, and graph-based reasoning, forming the foundation for richer, AI ready data discovery in SAP HANA Cloud. What used to take days of manual modeling can now be done automatically in minutes.
Create a custom database objects knowledge graph on SAP HANA Cloud Central
This Knowledge Graph allows agents to reason about data, to understand how entities connect, what relationships exist, and which attributes are relevant. It is the foundation that enables both discovery and querying to work contextually and at scale.
Custom Database Objects Knowledge Graph powers the SAP HANA Cloud Suite of Agents
How It Works
Imagine a business user who doesn’t know the database schema or table names, they simply have a question in mind and want to “talk to the database.”
For example, they might ask: “List all customers with their sales conversion probability and related details?”
Here’s what happens behind the scenes:
The Discovery Agent searches the Database Objects Knowledge Graph to identify which tables and views contain the relevant data.The Data Agent then uses that information to generate a query that retrieves the results without any manual lookup or SQL writing required.
In a few seconds, the user receives a clear answer. The process feels conversational, even though the agents are performing complex discovery and query generation under the hood.
Ask a Natural Language Query on your database objects, via Joule in SAP HANA Cloud Central
Explore insights built on the actual database objects and relationships in your SAP HANA Cloud instance.
Enables Data Discovery
With the Discovery and Data Agents, SAP HANA Cloud transforms metadata into a living semantic layer that spans its multi-model engines. The “agent users” can now query their data environment conversationally, explore hidden relationships, and even surface annotations and sample values to guide analytics. Together, these agents transform SAP HANA Cloud from a static data platform into a dynamic environment where users can reason, explore, and interact with data naturally.
Looking Ahead: Context and Memory for Agents (Product Vision)
The journey toward an agentic SAP HANA Cloud is just beginning. As these capabilities evolve, context and long-term memory will play a key role in shaping the next generation of agents. Future agents will not only discover and query data but also retain context from previous interactions, remembering what was asked, which datasets were used, and what outcomes were returned. Over time, this could enable more natural and connected conversations with data, where insights evolve instead of starting from scratch each time.
This concept of Agentic Memory will help create more continuous and connected user experiences, all while leveraging HANA Cloud’s trusted data management and governance foundation.
Conclusion:
Together, these agents make SAP HANA Cloud the foundation for a new generation of secure, context-aware, and agent-driven data experiences.
For agent users, this means:
A unified way to access data and metadata.Natural-language querying through familiar protocols (MCP).The potential for context persistence across multi-turn conversations.A path toward AI-grounded data exploration and graph-enhanced analytics.
With that, it’s certain: SAP HANA Cloud is THE database for SAP’s AI-native software architecture. For additional updates and announcements from SAP TechEd 2025, check @thomashammer ‘s blogpost here
The ChallengeEnterprises today work with massive volumes of data spread across tables, views, JSON collections, graphs, and text. Yet most users don’t know where that data lives or how to access it efficiently. Building a bridge between business questions and the underlying technical data model often requires deep schema knowledge and SQL expertise.At the same time, users expect to “talk to their data”, to search, ask, and get answers without navigating complex structures. But creating such experiences has often been challenging. Data is often siloed, metadata is hard to navigate, and the context needed to connect insights is missing.The Innovation: Discovery Agent and Data Agent in SAP HANA Cloud (planned QRC1 2026)At SAP TechEd 2025, SAP HANA Cloud takes a major step forward by becoming agentic and introducing a suite of agents designed to simplify how people discover and interact with data.In this blog, the focus will be on the Discovery Agent and the Data Agent.Together, they make it easier for agent users to find the right data objects, understand relationships, and generate the right queries, all without needing deep technical or SQL knowledge.SAP HANA Cloud Suite of AgentsDiscovery Agent: Exploring data context The Discovery Agent helps users find and understand the data available in SAP HANA Cloud. It explores the HANA catalog using a Knowledge Graph that exposes how objects relate to one another, making it easy to locate relevant datasets without manually browsing schemas. When users describe what they need, like for example, project details or customer transactions, the Discovery Agent can identify the right data sources based on context, relationships, and annotations. Under the hood, the agent generates SPARQL queries on the RDF graph of HANA objects, revealing how data structures connect.Data Agent: Turning intent into executable queriesOnce the relevant data objects are found, the Data Agent takes over.It translates natural language intent into executable queries, such as SQL. Users can simply describe what they want to retrieve, for instance, “show my current projects with partner involvement” and the Data Agent generates and executes the corresponding query.By combining discovery and query generation, these agents make data exploration seamless: the Discovery Agent provides context, while the Data Agent delivers answers.Powered by the SAP HANA Cloud Custom Database Objects Knowledge Graph (planned QRC1 2026)At the core of these agents lies the Custom Database Objects Knowledge Graph: an RDF graph automatically built from system catalog that captures tables, columns, relationships, and annotations. It provides the semantic layer needed for exploration, AI grounding, and graph-based reasoning, forming the foundation for richer, AI ready data discovery in SAP HANA Cloud. What used to take days of manual modeling can now be done automatically in minutes.Create a custom database objects knowledge graph on SAP HANA Cloud CentralThis Knowledge Graph allows agents to reason about data, to understand how entities connect, what relationships exist, and which attributes are relevant. It is the foundation that enables both discovery and querying to work contextually and at scale.Custom Database Objects Knowledge Graph powers the SAP HANA Cloud Suite of AgentsHow It WorksImagine a business user who doesn’t know the database schema or table names, they simply have a question in mind and want to “talk to the database.”For example, they might ask: “List all customers with their sales conversion probability and related details?”Here’s what happens behind the scenes:The Discovery Agent searches the Database Objects Knowledge Graph to identify which tables and views contain the relevant data.The Data Agent then uses that information to generate a query that retrieves the results without any manual lookup or SQL writing required.In a few seconds, the user receives a clear answer. The process feels conversational, even though the agents are performing complex discovery and query generation under the hood.Ask a Natural Language Query on your database objects, via Joule in SAP HANA Cloud CentralExplore insights built on the actual database objects and relationships in your SAP HANA Cloud instance. Enables Data Discovery With the Discovery and Data Agents, SAP HANA Cloud transforms metadata into a living semantic layer that spans its multi-model engines. The “agent users” can now query their data environment conversationally, explore hidden relationships, and even surface annotations and sample values to guide analytics. Together, these agents transform SAP HANA Cloud from a static data platform into a dynamic environment where users can reason, explore, and interact with data naturally.Looking Ahead: Context and Memory for Agents (Product Vision)The journey toward an agentic SAP HANA Cloud is just beginning. As these capabilities evolve, context and long-term memory will play a key role in shaping the next generation of agents. Future agents will not only discover and query data but also retain context from previous interactions, remembering what was asked, which datasets were used, and what outcomes were returned. Over time, this could enable more natural and connected conversations with data, where insights evolve instead of starting from scratch each time.This concept of Agentic Memory will help create more continuous and connected user experiences, all while leveraging HANA Cloud’s trusted data management and governance foundation.Conclusion:Together, these agents make SAP HANA Cloud the foundation for a new generation of secure, context-aware, and agent-driven data experiences.For agent users, this means:A unified way to access data and metadata.Natural-language querying through familiar protocols (MCP).The potential for context persistence across multi-turn conversations.A path toward AI-grounded data exploration and graph-enhanced analytics.With that, it’s certain: SAP HANA Cloud is THE database for SAP’s AI-native software architecture. For additional updates and announcements from SAP TechEd 2025, check @thomashammer ‘s blogpost here Read More Technology Blog Posts by SAP articles
#SAP
#SAPTechnologyblog
