What's New in EMQX Enterprise 5.10.0: AI-Powered Flows, NATS Gateway, and S3 Tables Integration

If you've been waiting to build smarter, more flexible data pipelines with MQTT, EMQX Enterprise 5.10.0 delivers.

This latest release introduces support for LLM-based message processing, natural language SQL rule generation, and a new NATS protocol gateway. It also adds expanded sink options for Apache Doris and Amazon S3 Tables.

Whether you're looking to integrate AI, connect across protocols, or build real-time analytics data flows, this update includes practical improvements that help you move faster and accomplish more.

AI-Powered Flows with LLM Support in Flow Designer

Flow Designer now supports LLM integration, allowing you to process MQTT messages using AI models like OpenAI and Anthropic Claude. With this update, you can create intelligent, low-code workflows that respond to incoming MQTT data using natural language prompts.

Key Benefits:

• Real-Time AI Enrichment Enrich or analyze MQTT messages in real time. For example, automatically classify sensor data using contextual AI insights.
• No-Code AI Pipelines Easily configure LLM nodes within Flow Designer's visual UI. These nodes use EMQX's ai_completion function to call external LLM APIs.
• Flexible Provider Support Connect to multiple AI providers and customize prompts, API keys, and settings.
• Optimized for Smart IoT Use Cases Ideal for scenarios like predictive maintenance, anomaly detection, or human-readable summaries when combined with third-party engine or logic, without writing code or SQL.

Example: In a smart city deployment, sensor data can pass through an LLM node to generate a human-readable summary like: “Air quality downtown is moderate, with elevated PM2.5 levels.” This allows decision-makers to act on insights instantly.

Learn more:

• LLM-Based Data Processing Overview

• OpenAI Node Quick Start

• Anthropic Node Quick Start

Write Rules Using Natural Language in the Rule Engine

Writing SQL rules just got easier. The new AI SQL Generator in the EMQX Dashboard allows users to describe data logic in plain language. EMQX then converts it into ready-to-use SQL statements for the Rule Engine.

Key Benefits:

• Natural Language to SQL Input: “Alert when temperature > 80” Output: Valid SQL rule syntax.
• Faster Rule Creation Speeds up rule configuration and lowers the technical barrier for new users.

Example Use Case

Task Description: A technician enters: “Trigger an alert if a message from any device under topic factory/+/status contains a temperature over 90 and a vibration level above 5.”

Related Topics: Specify a topic example such as factory/machine-42/status.

Input Example (MQTT Payload): Provide a sample MQTT message payload to help the AI understand your data structure.

{   
  "device_id": "machine-42",   
  "temperature": 95.6,   
  "vibration": 6.2,   
  "status": "running",   
  "timestamp": "2025-06-12T08:15:30Z" 
}

Output Example (optional): Specify the expected result format. Example:

{  
  "device_id": "machine-42",  
  "alert": "Temperature and vibration thresholds exceeded",  
  "temperature": 95.6,  
  "vibration": 6.2,  
  "timestamp": "2025-06-12T08:15:30Z"  
}

Note: This output is an example of an enriched alert message that could be generated by EMQX using a template in the rule action. It is not the raw result of the SQL query, but rather the expected format of a notification or alert sent when the rule condition is met.

The AI SQL Generator returns the following SQL rule:

SELECT *, payload.temperature, payload.vibration FROM "factory/+/status" WHERE payload.temperature > 90 AND payload.vibration > 5

Learn more about the AI SQL Generator.

New Data Integrations: Apache Doris & Amazon S3 Tables

Version 5.10.0 introduces two new data sinks for high-performance analytics and cost-effective long-term storage of MQTT data.

Stream MQTT Data to Apache Doris for Real-Time Analytics

EMQX now natively integrates with Apache Doris, a high-speed analytical database. Using the rule engine, MQTT messages are matched by topic, processed, and mapped into structured SQL templates. These messages are then written to Doris via HTTP or JDBC.

With this setup, you can query your IoT data in real time using standard SQL, and build live dashboards with BI tools like Grafana.

Apache Doris Integration Guide

Store MQTT Data in Amazon S3 Tables with Iceberg Format

EMQX also supports Amazon S3 Tables, using the Apache Iceberg format for large-scale data storage and analytics.

This integration transforms MQTT data into Iceberg-formatted tables and streams them directly into S3. It eliminates the need for traditional databases while retaining SQL-like querying capabilities.

Once stored, data can be queried through Amazon Athena, EMR, Redshift Spectrum, or third-party tools like Trino, Presto, and Snowflake.

Amazon S3 Tables Integration Guide

Bridge MQTT and NATS with New Protocol Gateway

EMQX now includes native gateway support for NATS, a lightweight and high-performance messaging protocol popular in microservices and cloud-native environments.

Key Features:

• MQTT-NATS Bridging Bi-directional data exchange via topic-subject mapping.
• Cross-Protocol Communication Messages published by NATS clients are translated into MQTT publishes and vice versa, including support for wildcards, queue groups, and request/reply patterns.
• Simple Setup Configure via Dashboard, REST API, or config files.

NATS Protocol Gateway Documentation

Simpler Rule Creation with Namespaced Event Topics

EMQX 5.10.0 introduces namespaced event topics for system events (e.g., client connections, disconnections), enabling easier rule creation with wildcard support.

Previously, users had to list each event topic explicitly:

SELECT * FROM "$events/client_connected", "$events/client_disconnected" WHERE clientid = 'c123'

Now, events follow a structured format like $events/client/connected, allowing simplified queries:

SELECT * FROM "$events/client/+" WHERE clientid = 'c123'

This makes it easier to match multiple related events (e.g., all client events) using a single wildcard pattern.

Refer to the Event Topic List for the supported event topics and the mapping between the old and new (namespaced) event topics.

Deeper Observability for MQTT Durable Sessions

New metrics at the API level improve monitoring for:

• RAFT replication and snapshot status: Track the log replication progress and snapshot generation status under the RAFT protocol.
• Shard Distribution and Performance: Visualize how session data is partitioned and distributed across cluster nodes, and assess the performance of each shard.
• Fault diagnosis: Identify potential issues such as slow or unstable nodes by observing replication delays and instability.

Other Improvements in Security, Performance, and UX

Simplify Access Control with LDAP-Based Rules

The LDAP authenticator now supports dynamic retrieval of ACL rules during authentication. MQTT topic-level permissions can be defined as attributes in the LDAP schema and cached per session. This unifies authentication and authorization processes.

Learn more about LDAP ACL Retrieval.

“Created At” Column in Namespace List

The Dashboard now shows creation timestamps for each namespace, enabling easier management and operational tracking in multi-tenant environments.

Kafka Connector: AWS IAM Authentication Support

Enables secure access to Amazon MSK using IAM roles, simplifying credential management.

Improved Clustering with Static Discovery (#15304)

EMQX 5.10.0 fixes an issue affecting replicant node discovery in clusters using the static discovery strategy. Previously, the replicants could ignore core nodes not explicitly listed in the static_seeds list. This could lead to an inconsistent cluster view and load imbalance.

With this fix, all core nodes are now correctly recognized by replicants, ensuring more consistent clustering and balanced resource usage in static-mode deployments.

This release also includes numerous performance improvements and bug fixes. For the full list of changes, see the EMQX Enterprise 5.10.0 Release Notes.

Ready to Explore EMQX 5.10.0?

Download EMQX Enterprise 5.10.0 from the official website and start building intelligent, scalable, and secure IoT solutions.

Need help or want to explore how EMQX fits your IoT use case? Contact our team for a personalized consultation.

Try EMQX Enterprise for Free

Get Started →