ASI Biont MQTT Integration: How to Control IoT Devices via Telegram Without Code

Introduction: Why MQTT and AI Are a Perfect Pair

If you work with IoT—temperature sensors, relays, smart lamps, or industrial controllers—you know that MQTT (Message Queuing Telemetry Transport) is the de facto standard for data exchange between devices. According to the Eclipse IoT Working Group (2025), over 70% of industrial IoT solutions use MQTT as the primary protocol. But classic MQTT work requires: setting up a broker (e.g., Mosquitto, HiveMQ, or EMQX), writing scripts to subscribe to topics, and manually processing data. This takes hours, especially if you manage dozens of sensors.

ASI Biont integration with MQTT changes the game. The AI agent connects to any MQTT broker via the platform's API connector—and you get the ability to control devices through Telegram in natural language. No code: you simply give the AI agent an API key or connection details to the broker, and it writes the integration layer itself. All interaction happens through a chat dialogue, without control panels or "add integration" buttons.

In this article, I'll show real examples of how to connect ASI Biont to an MQTT broker, automate temperature monitoring, control relays via Telegram, and save up to 80% of time setting up IoT infrastructure.

What Is MQTT and Why Connect It to an AI Agent

MQTT is a lightweight publish/subscribe messaging protocol. Devices (sensors, relays, controllers) publish data to topics (e.g., sensor/temperature), and other devices or services subscribe to these topics and receive updates. The broker (Mosquitto, EMQX, HiveMQ) acts as an intermediary.

Connecting MQTT to the ASI Biont AI agent allows:
- Real-time sensor data monitoring—AI interprets readings and provides recommendations.
- Device control via natural language—"Turn on the living room light" or "Set temperature to 22°C."
- Scenario automation—for example, if the temperature exceeds 30°C, AI sends a command to turn on a fan.
- IoT integration with other services—Telegram bot for control, Google Sheets for logging, Slack for notifications.

Without an AI agent, you'd have to write Python code (paho-mqtt library), set up webhooks, and manually process JSON messages. With ASI Biont, you simply describe the task in words.

How the Integration Works: From API Key to Ready Scenario

The main feature of ASI Biont is that it connects to any service via API, independently writing integration code for the specific protocol. For MQTT, this means:

1. You provide the AI agent with connection details—broker address (e.g., mqtt://broker.emqx.io:1883), login/password (if required), and a list of topics to subscribe to. All this is communicated in the chat in natural language: "Connect to my MQTT broker at mqtt://192.168.1.100:1883, topic sensor/temperature."

2. The AI agent generates integration code—it creates a Python script that connects to the broker via the paho-mqtt library, subscribes to the specified topics, and passes data to the ASI Biont core. You don't need to see this code—it runs in the background.

3. Scenario setup through dialogue—you say: "If the temperature in the sensor/temperature topic exceeds 28°C, send me a notification in Telegram and publish a command to the relay/control topic to turn on the fan." The AI creates the processing rules itself.

4. No control panels—the entire process from connection to automation setup happens in the chat. This is a key difference from traditional platforms (e.g., Node-RED or Home Assistant), where you need to manually configure nodes and write functions.

Example 1: Temperature Monitoring in a Greenhouse via Telegram

Imagine you have a DHT22 sensor connected to an ESP8266 that publishes temperature and humidity to the topic greenhouse/sensor. You want to receive Telegram notifications if the temperature goes outside the range of 18–30°C.

Without ASI Biont:
- Write a Python script to subscribe to the topic.
- Set up a Telegram bot (get a token, write a command handler).
- Deploy the script on a server or Raspberry Pi.
- Time: 2–3 hours, including debugging.

With ASI Biont:
You type in the chat: "Connect to the MQTT broker at mqtt://192.168.1.50:1883, subscribe to the topic greenhouse/sensor. If the temperature is above 30°C or below 18°C, send me a message in Telegram with the current readings." The AI agent creates the integration in 2 minutes, connects to the broker, and starts monitoring. A Telegram message arrives: "Greenhouse temperature: 32°C—exceeds norm. I recommend turning on ventilation."

Example 2: Controlling a Relay via Telegram with Voice

Suppose you have a relay connected to the topic home/light/living_room. The relay accepts commands on or off. You want to control the light via Telegram without complex interfaces.

Without ASI Biont:
- Set up an MQTT client on a server.
- Write a Telegram bot with buttons.
- Implement command parsing logic.
- Time: 4–5 hours.

With ASI Biont:
You say in the chat: "Connect to the MQTT broker, topic home/light/living_room. When I write 'turn on the living room light' or 'turn off the light,' send on/off commands to this topic." The AI agent sets up an NLP handler and integration in seconds. Now you can simply write: "Turn on the light"—and the relay activates.

Example 3: Smart Home Automation with Multiple Devices

In a real scenario, you might have 10+ sensors (temperature, motion, light) and 5+ relays (lights, air conditioner, ventilation). ASI Biont allows creating complex rules:

• "If the motion sensor in topic sensor/motion/parking detects activity after 10 PM, turn on the light for 5 minutes (topic relay/light/garage) and send me a notification."
• "If the temperature in topic sensor/temperature/bedroom is above 25°C and humidity is below 40%, turn on the air conditioner (topic relay/ac) and humidifier (topic relay/humidifier)."

All these rules are set up through dialogue. The AI agent subscribes to the necessary topics, parses JSON messages (e.g., {"temp": 23.5, "hum": 55}), analyzes data, and performs actions.

Benefits of Integration: Time and Money Savings

According to a McKinsey study (2024), automating IoT operations with AI reduces operational overhead by 40–60%. In my case, after connecting 15 sensors to ASI Biont, I stopped spending weekends debugging MQTT scripts—everything works through Telegram.

Step-by-Step Guide: How to Connect MQTT to ASI Biont

1. Go to the chat with the AI agent on the asibiont.com platform. No control panels—just a dialogue.
2. Enter a connection command, for example: "Connect to the MQTT broker at mqtt://test.mosquitto.org:1883, subscribe to the topic test/temperature." If the broker requires authorization, add login and password: "...with login admin and password pass123."
3. Wait for confirmation. The AI agent will report that the integration is active and start listening to topics.
4. Set up scenarios. Write: "If the value in the test/temperature topic is above 30, send me a message in Telegram." The AI creates the rule.
5. Control devices. Commands like "Publish the command on to the test/relay topic" will work instantly.

Why This Is Secure

ASI Biont uses encryption for data transmission (TLS) and does not store API keys in plain text—they are encrypted on the platform side. For MQTT brokers with TLS support (e.g., HiveMQ Cloud), the AI agent automatically establishes a secure connection. You can also restrict broker access by IP addresses—ACL (Access Control Lists) support is configured through dialogue.

Conclusion

ASI Biont integration with MQTT is not just a connection—it's full automation of IoT device management. You eliminate the need to write scripts, set up servers, and debug errors. All you need is to describe the task in words. Time savings: from 4 hours to 2 minutes per device. Money savings: lower development and maintenance costs.

Try it yourself—connect your MQTT broker to ASI Biont right now at asibiont.com. Just write in the chat: "Connect MQTT broker"—and the AI agent will do the rest.