Breaking the Batch Barrier: How Confluent Unlocks Real-Time SAP Innovation

SAP systems have long supported global organizations, running core areas like finance, supply chain, manufacturing, and human resources. Their reliability has made them a trusted foundation for decades.

What’s changing is how SAP data is used.

Increasingly, SAP is not only a system of record but also a system of insight—one that depends on timely data from across an enterprise. Customer interactions, Internet of Things (IoT) signals, ecommerce activity, partner systems, third-party data sources, and manufacturing and inventory systems all play a role in enriching core enterprise resource planning (ERP) data. When that data flows into SAP in real time, it enhances analytics, planning, and operational decision-making without disrupting the integrity of the core system.

Batch updates, though reliable, can slow down decision-making and limit flexibility. Real-time data flow, on the other hand, allows businesses to respond faster, improve efficiency, and create better customer experiences. 

Apache Kafka® makes this possible by streaming events into SAP landscapes. Confluent’s data streaming platform, with unified Apache Kafka and Apache Flink®, integrates bidirectionally with SAP Business Data Cloud (BDC), delivering real-time access to hundreds of pre-built data sources to fuel AI-powered digital experiences and business intelligence (BI). In this post, we’ll explore how Confluent helps organizations build comprehensive data products for SAP BDC. 

Ready to learn common streaming use cases you can implement when setting your ERP data in motion with Confluent and SAP? 

The Need for Better SAP Integrations

As enterprises have embraced digital transformation, the role of SAP data has evolved beyond powering internal transactions. Businesses now expect SAP to integrate seamlessly not just within SAP’s own ecosystem but also across the entire enterprise data landscape: modern cloud applications, analytics platforms, data lakes, customer relationship management (CRM) systems, and artificial intelligence (AI) models. 

However, most existing SAP integrations were designed in a different era, when they were optimized for stability at the cost of speed, flexibility, and data accessibility.

To truly unlock innovation, organizations need a modern integration approach that lets SAP data flow continuously and contextually across systems. This means moving from rigid, batch-based transfers to agile, event-driven architectures that can keep up with today’s real-time business demands. The following sections explore the key challenges of traditional SAP integrations and why Confluent’s data streaming platform, in partnership with SAP, is reshaping how enterprises connect, share, and act on data.

Too Many Connections

Over the years, SAP has been connected to other systems in almost every way imaginable: direct database links, custom code, ETL jobs, and enterprise service buses (ESBs) such as TIBCO, SOAP, or REST APIs. Each approach worked for its time, but the result was often a patchwork of integrations that were complex and hard to evolve. Today’s enterprises need something more flexible, loosely coupled, and easier to adapt as business needs change.

This is one reason that SAP and Confluent have partnered to simplify and modernize how data moves across enterprise systems. Through this collaboration, Confluent provides a data streaming platform that integrates bidirectionally with SAP Datasphere within the SAP BDC suite, enabling customers to connect SAP and non-SAP systems with a single, consistent, and governed data layer.

Batch Delays Data Insights

Traditionally, SAP data moves in scheduled batches every 12 hours, nightly and weekly. That pace made sense when businesses needed only daily or monthly reports. But today, requirements have shifted:

• Dashboards need real-time updates.

• Systems must react instantly to new orders and events.

• Batch jobs add unnecessary coordination overhead.

• AI models and modern automations work best with live, continuous data.

Old Middleware Is Heavy

Many organizations still rely on traditional middleware such as IBM MQ, TIBCO Enterprise Message Service (EMS), and other license-heavy platforms to move SAP data. These tools were reliable for decades, but real-time streaming calls for lighter, more scalable options that can handle critical workloads without the same complexity or cost.

Mixing Data Opens New Possibilities

The real opportunity comes when SAP data combines seamlessly with other sources, IoT devices, CRM streams, supply chain partner updates, and more. While batch ETL can make this possible, event streaming delivers it continuously, opening the door to faster insights and smarter automation.

Custom Code Creates Maintenance Work

Custom glue code and point-to-point integrations gave teams full control, but they also created long-term maintenance effort. Many businesses now prefer reusable, managed, standards-based integrations that reduce custom development.

Modern enterprises are moving toward a more connected data landscape where SAP systems don’t operate in isolation but function as part of a live, event-driven ecosystem. This is where Confluent comes in to provide the real-time streaming foundation that helps SAP stream data seamlessly throughout a business.

How Confluent Connects SAP Data in Real Time

Confluent delivers a modern streaming backbone that complements your SAP systems with continuous, flexible data sharing. Instead of waiting for scheduled batches, Confluent allows SAP data to flow as live event streams, enabling organizations to see, react, and decide instantly.

Live Streaming Instead of Batch

With Confluent, SAP data flows as live event streams—no more waiting for the next batch.

Dashboards update live. Fraud alerts trigger instantly. Production and inventory signals move across sites and partners with minimal delay. This real-time movement turns static business processes into living, connected data streams that fuel faster decisions and higher efficiency.

Flexible, Loosely Coupled Integration

Kafka’s log-based, distributed design decouples data producers and consumers, allowing each side to evolve independently. In SAP integrations, this means data can flow in both directions. SAP systems can act as producers that stream events and transactions into Kafka or consumers that receive updates and insights from other systems.

This loose coupling gives enterprises flexibility to extend their SAP landscapes without brittle point-to-point links. 

Confluent also adds enterprise-grade capabilities on top of Kafka, including monitoring, security, schema management, and operational visibility, so your SAP data remains consistent, reliable, and governed across the entire ecosystem.

Whether you’re streaming orders out of SAP ERP Central Component (ECC) or pushing enriched AI predictions back into SAP S/4HANA, Confluent provides a scalable, secure, and governed data backbone to make it seamless.

Enabling Modern Real-Time Use Cases

It’s not just faster for the sake of it. Real-time streaming unlocks real impact:

• Automotive manufacturers monitor product quality and warranty signals in real time to identify and resolve issues proactively.

• Retailers deliver live operational dashboards fed directly from SAP transactions and maintain real-time inventory levels to support seamless customer experiences across stores and digital channels, along with smarter replenishment.

• Pharmaceutical manufacturers track manufacturing processes continuously, optimizing production speed and reducing costs to meet market demand.

• Financial services spot fraud patterns instantly and launch automated workflows immediately.

Confluent turns SAP’s core processes into a continuous stream of intelligence that drives action across systems, teams, and partners in real time.

Integrating Confluent With SAP BDC

Not every business process needs real-time updates, but every business needs connected, trusted data. That’s where Confluent’s integration with SAP BDC comes in.

Confluent acts as the connective tissue that brings together SAP and non-SAP data into SAP Datasphere, ensuring that information from across a business—finance, sales, supply chain, CRM, IoT, and more—can be accessed, enriched, and analyzed in one place.

Building Comprehensive Data Products With Confluent and SAP

SAP BDC provides a library of pre-built, semantically rich data products representing core business domains such as finance, inventory, orders, customers, and supply chain processes. These SAP-delivered data products form the foundation of the business data fabric and power SAP’s analytics, planning, and Business AI applications.

Through Confluent’s bidirectional integration with SAP Datasphere, organizations can expand these standard SAP data products with real-time events from across the enterprise IoT sensor readings, point-of-sale transactions, CRM activities, market feeds, clickstreams, and other operational data.

This continuous enrichment enables customers to build comprehensive data products that reflect the full operational reality of the business, not just what lives in SAP.

These enriched, real-time data products then power SAP’s Insight Apps, 360° analytics, extended planning, and Business AI with live, complete, and trusted information. In effect, Confluent transforms SAP’s traditionally batch-driven ecosystem into a dynamic, real-time data foundation capable of supporting predictive, proactive, and AI-driven business outcomes.

[Caption] How Confluent Enriches SAP Data in Real-Time for Applications, Data Products, and AI

To deliver these comprehensive, real-time data products, Confluent provides the streaming infrastructure that continuously synchronizes SAP and non-SAP systems. The following capabilities explain how Confluent connects data with SAP Datasphere to keep business information fresh, complete, and ready for analytics, planning, and AI.

Bidirectional Real-Time Data Integration

Confluent supports seamless, two-way streaming of data between SAP systems such as S/4HANA or ECC and SAP Datasphere. This keeps ERP and operational data continuously synchronized across SAP and non-SAP environments, ensuring that business insights and analytics always run on the latest information.

Streaming Non-SAP Data into SAP Datasphere

Through Confluent’s extensive portfolio of pre-built connectors, organizations can stream live data from external source databases, software-as-a-service (SaaS) applications, cloud platforms, and IoT devices into SAP Datasphere. This allows analytics, planning, and AI tools within SAP to combine ERP data with real-time inputs from logistics, customer systems, or connected equipment, delivering faster and richer insights.

Turning Raw Data into Smart Insights

As data moves through Confluent, it can be enriched with business context from SAP services, turning raw, transactional data into meaningful, structured information. This semantic enrichment helps users perform deeper, real-time analysis that reflects true business processes rather than isolated data points.

When this enriched data powers AI and machine learning models, companies can detect anomalies, predict demand, optimize supply chains, and personalize customer experiences with fresh, contextual information.

Governance, Security, and Simplified Operations

Confluent provides enterprise-grade governance through schema management, role-based access control (RBAC), and end-to-end monitoring, ensuring that all data entering SAP Datasphere remains consistent, trusted, and secure. While governance and schema management can sound abstract, its value becomes very concrete in real-world SAP landscapes.

For example, when a SAP team updates an Operational Data Provisioning (ODP) extractor or modifies a Core Data Services (CDS) view adding a field, renaming a column, or changing a data type, Confluent Schema Registry automatically detects the schema change and validates it against downstream contracts. This prevents data lake pipelines, BI dashboards, or analytics workloads from silently breaking due to unexpected structural changes. Instead of spending hours debugging incompatible fields, teams receive early warnings and compatibility checks that keep the entire data ecosystem stable.

Schema Registry also ensures predictable, versioned data delivery for all downstream consumers. This gives analytics, AI, and data engineering teams the confidence to build on SAP data streams without fear of instability when SAP evolves. Combined with RBAC and centralized monitoring, Confluent delivers practical, automated governance that aligns with SAP’s compliance and data quality standards.

By replacing batch-based middleware and reducing custom integrations, Confluent modernizes how SAP Datasphere connects with the rest of an enterprise. The result is lower maintenance overhead, faster cloud migrations, and smoother interoperability across systems.

Example Scenarios

• Streaming SAP ERP transactions directly into SAP Datasphere for live operational dashboards

• Identifying manufacturing bottlenecks, cutting lead times, and enabling predictive maintenance with real-time SAP and IoT data

• Enriching SAP financial data with CRM or IoT feeds for unified business analytics

• Delivering SAP Datasphere data to downstream platforms such as Databricks, Snowflake, or Amazon Redshift in real time

• Powering risk management, personalization, and predictive maintenance with live, contextual SAP data

Together, Confluent and SAP BDC eliminate the divide between operational and analytical systems. They create a live, governed, and intelligent data backbone where every part of the business, from finance and supply chain to AI and analytics, runs on the same, up-to-date information.

It’s not just about making data real time. It’s about making it connected, consistent, and ready to drive value wherever and whenever the business needs it.

6 Ways to Connect SAP and Apache Kafka® Workloads

SAP has powered the beating heart of global business for decades. It’s stable, deeply integrated, full of history, and full of surprises. But as more companies build modern data stacks around Kafka, there’s one question everyone runs into sooner or later: How do we actually connect our trusted, reliable SAP to our real-time pipelines?

Before we dive into all the do-it-yourself patterns, it’s worth calling out the fully managed, productized path that many customers start with: SAP Datasphere/SAP BDC with the Confluent integration. 

SAP ships native replication flows in Datasphere that can stream SAP data (for example, CDS views and local Datasphere tables) directly into Confluent and increasingly read from Confluent as a source using a SAP-built Kafka connector that’s available and supported as a SAP Endorsed App via the SAP Store. If you’re already standardizing on Datasphere/BDC as your business data fabric, that will likely be the best option to first evaluate, and we’ll come back to it explicitly in Option 5.

There isn’t just one option because: 

• Sometimes you’re not using Datasphere/BDC at all and still need to connect SAP and Kafka.

• Sometimes you’re using Datasphere/BDC but need to extend it with additional patterns (for example, for non-SAP systems, specialized change data capture [CDC] or stricter latency/ops needs) beyond the out‑of‑the‑box integration.

The six options below are best thought of as complementary patterns around that productized path, and the right one depends on where your data sits, how fresh you need it, who’s allowed to touch what, and how much you trust your own team to maintain moving parts. So let’s break down the real options.

Option 1: Database-Centric Connectors

One of the oldest tricks in the book is to pull directly from SAP’s database. Many teams stick with the tried-and-true JDBC approach, especially when they need raw replication fast.

• Confluent JDBC Connector: One of the most straightforward ways to get data out of SAP and into Kafka is still the classic—read straight from the database. Using Confluent’s JDBC Connector with the official SAP HANA JDBC driver (ngdbc.jar), teams poll HANA tables on a schedule and push rows into Kafka topics with no Advanced Business Application Programming (ABAP) and no NetWeaver tweaks. 

It’s fully supported by Confluent, quick to spin up, and requires no ABAP work or NetWeaver tweaks. However, it gives you only raw tables—no SAP business logic, no deltas—so you’ll need to stitch the business meaning back together yourself.

• SAP HANA Source Connector: A partner-built alternative that works similarly to JDBC, with some extra HANA-specific tweaks like offset tracking. But watch out: Some of these are not actively maintained or Confluent-supported.

It’s helpful if your workloads are HANA-heavy and you want more control. However, some partner HANA connectors aren’t actively maintained or officially supported by Confluent anymore, so vet carefully if long-term stability and vendor support are important.

Option 2: ODP Extractors

SAP’s ODP is purpose-built for sending out deltas, not brute-force full table pulls. When you want business context, ODP is a strong bet.

INIT Connect for SAP ODP Source: This Kafka Connect–native connector hooks straight into SAP extractors (NetWeaver, ECC, BW/4HANA) and then streams those business-level deltas into Kafka, fully in sync with SAP’s logic. It uses JCo and Remote Function Call (RFC) to talk to SAP and streams only the deltas you actually care about.

ODP preserves SAP’s logic; you don’t have to rebuild joins in Kafka later. However, the SAP team needs to configure and maintain those extractors, which can break when you upgrade SAP or make schema tweaks. It’s not “fire and forget”; someone has to look after it.

The blog post Apache Kafka and SAP Integration With the Kafka Connect ODP Source Connector explores how enterprises can achieve scalable, near–real-time SAP data integration with Apache Kafka using Kafka Connect, INIT Software’s ODP Source Connector for ABAP-based SAP systems.

Option 3: Log-Based CDC

Sometimes the fastest way to see what changed is to watch the database (DB) transaction logs directly. CDC skips the SAP layer altogether.

• Qlik Replicate: This is a well-known tool for log-based CDC that reads Oracle, IBM Db2, or HANA logs and pushes row changes into Kafka.

It’s well suited for real-time replication of SAP database changes without needing to set up SAP extractors or touch ABAP. However, it stays read-only, so you can’t push updates back—and like Qlik, it doesn’t support HANA log capture or CDS views out of the box. So if you’re planning to modernize your SAP core, check compatibility early and expect some limits for newer HANA-backed workloads.

• Precisely Connect: This is another CDC option for the same job. It taps your database logs and streams inserts, updates, and deletes in near–real-time, mainly for IBM Db2 and Oracle under older ECC environments. 

It’s ideal for teams that want straightforward, continuous replication of SAP DB changes to Kafka with minimal SAP-side setup. However, it’s strictly read-only, so you can’t push anything back and there’s no support for HANA or modern CDS views. If you’re planning a shift to HANA, you’ll probably need another pattern later.

Option 4: Application and API Connectors (Events, OData, Web Service)

This is where real time gets smart. Instead of pulling raw rows, you subscribe to business events or pull clean objects via APIs or push data back into SAP.

Four solid options stand out:

• SAP Integration Suite Kafka Adapter. SAP itself offers a certified Kafka Adapter inside its Integration Suite that can be used to fetch records from or send records to Kafka topics. It’s cloud-first but works on-premises too. If you already run Cloud Platform Integration (CPI), this is the official route source and sink flow supported; SAP’s own team maintains it. It pushes out every update.

• ASAPIO Integration Add-On: This is a smart NetWeaver add-on connector for getting data out of SAP systems into Confluent Cloud or Confluent Platform. It sits inside your SAP stack, listens for transaction events, enriches the data, and then streams it to Kafka via JSON/REST Proxy. Again, this is source and sink and hands-off once installed; SAP teams manage it. Every update is captured.

• INIT Connect for SAP Business Events Source: This is a Kafka Connect–native setup that taps into SAP’s internal push events and then pulls extra details using OData when needed. It runs as a source connector for the SAP Business Events Subscription OData API and is built on the Kafka Connect framework, so it automatically supports pluggable converters such as Apache AvroTM or JSON, single message transforms, graceful back-off, and other helpful features. It’s fully wired into Kafka and ops, with built-in support for single message transforms (SMTs), data conversions, and Schema Registry. It’s commercial-ready and managed operationally by Confluent. The only thing to watch out for is timing: There can be a gap between when SAP fires a change event and when the OData call picks up the rest of the data.

• INIT Connect for SAP OData v2: It connects directly to SAP’s OData v2 API on NetWeaver and works as both a source and a sink. Because OData is a web service, it polls SAP business objects to pull data into Kafka or push updates back. Built on the Kafka Connect framework, it automatically supports pluggable converters such as Avro or JSON, single message transforms, graceful back-off, and other handy features. It’s fully integrated with Kafka and ops, supports SMTs, conversions, and Schema Registry and is managed by the Kafka team.

• INIT Connect for SAP Web Service Data Sources Sink: A Kafka Connect–native setup that pushes Kafka data into SAP BI web service data sources. It works as a sink and is fully integrated with Kafka and ops, with built-in support for SMTs, data conversions, and Schema Registry. Built on the Kafka Connect framework, it supports pluggable converters such as Avro or JSON, single message transforms, and graceful back-off. It’s commercial-ready and managed operationally by the Kafka team.

Option 5: Productized Datasphere/BDC + Batch and Analytics Loaders

Sometimes for large-scale analytics or warehousing, you may be fine with daily or hourly loads. This is where SAP Datasphere or the SAP OpenHub Service comes in.

The productized path: SAP Datasphere/BDC + Confluent If you’re already investing in SAP Datasphere, the “happy path” is the native, jointly productized integration between Datasphere replication flows and Confluent. SAP ships a Kafka/Confluent connection inside Datasphere that can:

• Replicate SAP data (for example, S/4HANA CDS views or local Datasphere tables) into Confluent topics via replication flows

• Treat Confluent as a streaming endpoint for building real-time data products that combine SAP and non-SAP data

Confluent Cloud is listed in the SAP Store as a SAP Endorsed App for Datasphere, so this route gives you a shrink-wrapped, co-supported way to hydrate Datasphere with governed SAP business data and then fan out that data to warehouses, lakehouses, AI platforms, and operational apps without custom glue.

Classic batch and OpenHub-style loaders Beyond that productized Datasphere/BDC path, both Datasphere and SAP OpenHub use SAP’s ODP/CDS layers to deliver structured data on a schedule. They’re reliable, mature, and perfect for big data lakes, but you won’t get minute-level freshness. You fall back to more traditional “batch and analytics loader” patterns when:

• You’re constrained to existing BW/OpenHub investments.

• Your use case is strictly analytical and doesn’t justify standing up a full real-time streaming path yet.

Option 6: Custom Glue, SOAP, and REST

When all else fails, teams build what they need themselves. SAP’s software development kits (SDKs), like the NetWeaver RFC SDK or SAP JCo, let you wire up custom calls in Java, C++, or Node. You can also push events using ABAP Transmission Control Protocol (TCP) channels or work with SAP Cloud Platform Enterprise Messaging, which uses Solace under the hood to support Advanced Message Queuing Protocol (AMQP) or Java Message Service (JMS).

And then there’s good old SOAP and REST. Many older SAP modules only expose SOAP/Web Services Description Language (WSDL), so you wrap it with the Confluent REST Proxy or your own producer scripts. It works, but you own the complexity and the maintenance.

What’s the “Right” Way to Connect SAP and Kafka?

Researching and deciding which option is right for you is not an easy task, given the variety of integration alternatives, SAP products, interfaces, and use cases. If you want to explore the full landscape, the blog post Kafka SAP Integration – APIs, Tools, Connector, ERP, et al. breaks down different options, use cases, and trade-offs.

A key choice is whether you need an external third-party solution (like an ESB or ETL) or a Kafka-native integration (often built on Kafka Connect). Both have pros and cons. If you’re deciding between these paths, Apache Kafka vs Enterprise Service Bus (ESB) – Friends, Enemies, or Frenemies? dives deeper into how ESBs, ETL tools, and Kafka Connect differ in concept and architecture.

Ultimately, there’s no single “best” connector for SAP and Kafka, and there never will be. Different pieces solve different gaps, and most teams end up using more than one.

• Use JDBC or CDC when you just need raw tables moving quickly and don’t want to wait on SAP changes.

• ODP is better when you want proper deltas and would rather let SAP handle the business logic.

• Business Events or OData is useful when you care about actual structure and near–real-time data that makes sense out of the box.

• Web service sink lets you push data back into SAP instead of just pulling it out.

• Glue code or SOAP still happens when you have no other option and something custom needs to bridge the gap.

• Official SAP Kafka Adapter or ASAPIO is worth it if you want your SAP team fully on board and would rather keep it all blessed and supported.

The real trick is knowing where each piece fits, who owns it, how it might break, and how to keep the whole setup clean when SAP changes under the hood. Get that right, and you don’t just break the batch barrier; you turn SAP’s steady backbone into a live stream that keeps the rest of your data moving in real time.

The Takeaway

Your SAP systems are reliable, powerful, and at the heart of your business—but they weren’t built for the real-time world we live in today. With Confluent, you don’t have to replace SAP; you can simply unlock its full potential. Now, what happens in finance, production, or supply chain can flow instantly—not hours later—to every system that needs it.

Batch processes worked in the past, but today, they slow things down when customers and teams expect answers right away. A streaming data backbone keeps your SAP data live and moving to help your business make faster, smarter decisions.

By combining Confluent with SAP BDC, organizations can finally close the gap between day-to-day operations and real-time intelligence. SAP Datasphere keeps data trusted and meaningful while Confluent streams every transaction and event as it happens.

Together, they turn SAP from a traditional system of record into a system of real-time insight where analytics, AI, and business processes all run on live, accurate, and governed data.

Apache®, Apache Kafka®, Kafka®, Apache Flink®, Flink®, Apache AvroTM, and AvroTM are registered trademarks of the Apache Software Foundation. No endorsement by the Apache Software Foundation is implied by the use of these marks.