This release takes MQTT a step further by adding native support for the A2A protocol over MQTT, enabling AI agents to register, discover, and collaborate directly through the broker without additional infrastructure.
This article attempts to break down the full chain of IoT device intelligence from an engineering management perspective, analyze the root causes of efficiency bottlenecks, and explore a potential path forward.
Explore how A2A over MQTT enables cross-runtime AI agent discovery, communication, load balancing, and task coordination, with EMQX 6.2 bringing production-ready agent registry capabilities.
Learn how AI agents and MQTT power a decentralized IoT architecture, enabling per-device intelligence, self-coordinating workflows, and scalable device management with EMQX.
Explore how EMQX enables scalable Agent-to-Agent (A2A) communication with MQTT, powering event-driven AI agent orchestration in smart factories and IoT.
Discover how EMQX Neuron leverages edge computing, data aggregation, and contextual intelligence to eliminate the Cloud Ingress Tax, dramatically reducing IoT data costs, cloud compute expenses, and network overhead for Industrial AI deployments.
Learn how MQTT and EMQX power real-time, event-driven connectivity between AI and the physical world.
Discover how EMQX Unified Namespace empowers AI transformation in automotive manufacturing, enabling predictive maintenance, quality control, and human-robot collaboration across smart factories.
Learn how to give your AI companion visual perception using ESP32, MCP over MQTT, and multimodal LLMs for natural, human-like interactions.
Learn how to give your ESP32 AI companion memory, emotion, and persona over MQTT, transforming it from a simple controller into a personalized life assistant.
This is Part 4 of our Building AI Companion with ESP32 & MCP over MQTT series, introducing how to implement ASR and TTS for voice interaction on IoT devices.
In this article, we will build on that foundation by having a large language model (LLM) take over the control logic, creating a natural language command-and-response loop.
In this article, we'll extend the MQTT foundation from the previous post into an MCP over MQTT architecture.