Most organizations don’t struggle with real-time analytics because of dashboards or models. They struggle because data arrives too late to matter.

My learning is that the fastest teams aren’t just rebuilding pipelines or rewriting ingestion frameworks. They’re connecting what they already have—securely and at scale—and turning signals into decisions in seconds.

After my recent article Microsoft Fabric RTI: From Dashboards to Decisions, many of you asked the same question:

How can we accelerate time-to-market using existing data sources BUT without rebuilding pipelines or writing tons of code?

The answer, in many cases, is simpler than expected: Eventhouse connectors in Microsoft Fabric.

Eventhouse Connectors: More Than Just Ingestion

In real-time architectures, ingestion is not a background concern. It directly determines:

• How quickly signals become queryable?

• How flexible your RTI architecture can be?

• How well you manage schema evolution, throughput, and latency?

For sure, Microsoft Fabric is designed to meet organizations where they are—technically and operationally.

For organizations with Azure-centric environments, Eventhouse provides native, fully managed integration with services such as Azure Event Hubs, Azure Functions, Azure Data Factory, Power Automate, and Azure Stream Analytics. Also, Fabric connects directly to widely adopted open-source technologies including Apache Kafka and Flink, OpenTelemetry, Fluent Bit, Spark, Logstash, Telegraf, NLog, Serilog, and Splunk forwarders.

The diagram above illustrates how Microsoft Fabric RTI brings together first-party Azure services and open-source or third-party ingestion tools into a unified Eventstream pipeline. This pipeline delivers value-ready data directly into an Eventhouse KQL database, where logs, metrics, telemetry, and business events converge within seconds— and ready for KQL analysis!! —without requiring any re-architecture of existing systems.

To guide architectural decisions, I use the table below to compare the ingestion technologies supported by Fabric RTI, highlighting their market adoption, core capabilities, streaming support, and relative setup complexity based on my observations.

KQL databases in an Eventhouse natively store data in a format optimized for real-time analytics.

I want to highlight Apache Kafka because, in my experience, it unlocks real-time intelligence for many organizations. For example, events can come from Oracle CDC (Change Data Capture). Using the Kafka endpoint from an Eventstream custom endpoint source, this data can be streamed into Microsoft Fabric EventHouse for real-time related scenarios. If your data is in Oracle Cloud Infrastructure (OCI), you can use Kafka Connect with OCI Streaming. Kafka connectors are available for key Oracle services, including Oracle Object Storage (via S3), Oracle Integration Cloud, Oracle Database (JDBC), and Oracle GoldenGate (to send data from Autonomous DB and Siebel CRM - as an example).

As a reminder, to allow other Microsoft Fabric items—such as Lakehouses, Warehouses, Notebooks, Dataflows, or Reports—to query streaming data stored in a KQL database, the database must be connected to OneLake. Once connected, the Eventhouse truly becomes the illustrated gold mine: data is stored in a single, common format using Delta Parquet, an open standard.

Data in OneLake is automatically indexed for discovery and enriched with enterprise capabilities, including Microsoft Information Protection (MIP) labels, lineage, PII scanning, sharing, governance, and compliance. This is the right architectural choice.

By connecting the KQL database to OneLake, Fabric automatically exposes KQL tables as Delta Lake tables in OneLake. This enables other Fabric engines—such as Spark, SQL, Lakehouse, Warehouse, Dataflows, and Semantic Models—to query the same data directly through OneLake, fulfilling the requirement for unified, real-time analytics.

Low-Code and No-Code Ingestion Patterns in Practice

Because these connectors are native to Microsoft Fabric, many real-time pipelines require little to no custom code. To do so, we can use Get Data in RTI Eventhouse as it offers a step-by-step process to guide us through importing or inspecting the incoming data, creating, or editing the destination table schema, to exploration of the ingested result from multiple sources.

This is just one approach—other options exist depending on our scenario:

• Power Automate can capture SaaS events —such as SharePoint or Onedrive changes, Microsoft Dynamics 365 updates, or Teams activity— and route them through Eventstream directly into Eventhouse, all without scripting.

• Azure Data Factory can ingest logs, JSON payloads, CDC feeds, or API responses and write them directly into Eventhouse using RTI connectors, eliminating the need for custom ETL frameworks.

• If telemetry is already flowing through Event Hubs, Eventstream can be connected directly, routing data into KQL tables with no changes to the source system.

• For infrastructure and platform teams, lightweight agents such as Fluent Bit or OpenTelemetry Collectors can stream logs, metrics, and traces into Fabric with minimal configuration—often without modifying the application itself.

Even advanced scenarios, such as Spark Structured Streaming, can push data directly into Fabric RTI endpoints with a simple write operation, avoiding unnecessary middle layers.

Use case to showcase RTI in Cement Delivery Operations

Let’s consider a hypothetical scenario to illustrate the solution. A regional cement manufacturer, operating multiple plants and distribution hubs, was facing recurring challenges with on-time delivery to construction sites. While production capacity was strong, limited real-time visibility across dispatch, fleet operations, and site deliveries meant delays were often discovered too late—after crews were idle and costly penalties had already been incurred.

Leadership set a clear objective: gain real-time visibility into production, loading, and delivery execution without disrupting existing operational systems.

Using Microsoft Fabric RTI, the company connected live signals from plant systems, dispatch platforms, weighbridges, GPS-enabled truck fleets, and order management applications into a single, secure analytics layer. Existing data sources were integrated as-is, avoiding costly pipeline rewrites or operational downtime.

Within seconds of events occurring, operational data became queryable and actionable. Dispatch teams detected loading delays as they happened. In this example, logistics managers could track their fleet movement against delivery commitments in real time. Exceptions—such as traffic disruptions or unloading bottlenecks at construction sites—were identified early, enabling proactive intervention.

Security and governance were built in from day one. All ingestion was authenticated using enterprise identity, network access was tightly controlled, and operational telemetry inherited the same governance policies applied to business data. This allowed insights to be shared confidently across operations, logistics, and executive teams.

The impact was measurable: improved on-time-in-full delivery, reduced idle time at construction sites, increased fleet utilization, and stronger customer trust.

For leadership, the conclusion was clear: real-time intelligence became a strategic operational capability, not just an IT initiative.

Why This Matters

With Eventhouse, organizations don’t need to rewrite ingestion architectures to achieve real-time insight. They connect what they already have—Azure-native, open-source, or third-party systems—into a single RTI platform that delivers:

• Faster time-to-insight

• Lower operational overhead

• Enterprise-grade security and governance

• Minimal code with maximum interoperability

In industries where timing defines outcomes—manufacturing, logistics, oil & gas, energy, or retail—secure real-time intelligence is no longer optional. It’s a competitive advantage.

To learn more, please review the official documentation: Overview of data connectors - Microsoft Fabric | Microsoft Learn