Voice-Activated AI Task Planner for Telegram and Notion - n8n Workflow

Productivity Enthusiasts: Individuals seeking frictionless task capture via voice or text messaging.
AI Power Users: Users who want to leverage advanced LLM capabilities (like GPT-4o) for semantic search and complex task analysis.
Technical Teams: Groups needing robust n8n templates for integrating communication apps (Telegram) with project management systems (Notion).
n8n Developers: Those looking for a complex, production-ready n8n node chain demonstrating conditional routing and AI parsing.

Managing tasks often involves friction, especially when ideas strike on the go. This advanced n8n workflow eliminates that friction by allowing users to simply send a voice note or text message to a dedicated Telegram bot.

The central logic of this n8n automation uses AI (GPT-4o) to instantly detect the user's intent: Are they trying to create a new task, update an existing one, or analyze their current workload? This classification drives the subsequent process.

For example, if you say, “Add a task to call the dentist next Tuesday,” the n8n node sequence automatically transcribes the audio (if needed), extracts the title, due date, and priority, and creates the page in Notion. If you ask, “Am I too busy this week?”, the workflow analyzes all your current Notion tasks and provides a thoughtful summary back via Telegram. This is a highly practical and sophisticated example of an n8n workflow built for daily productivity.

This comprehensive n8n workflow is triggered by the Receive Telegram Messages n8n trigger.

1. Input Handling: The flow first determines if the input is Voice or Text using an n8n Switch node. If it's a voice note, the Telegram n8n node fetches the file, and an OpenAI n8n node transcribes it. Text messages skip this step.


2. Intent Detection: The unified text payload enters the core AI processing stage. A ChainLLM n8n node, powered by GPT-4o-mini, is used to classify the user's command into one of three intents: create, update, or analyze.


3. Conditional Routing: An n8n Switch node (Route by Intent) sends the execution down one of three specialized paths:

1. Confirmation: In all successful paths (Create, Update, Analyze), a final Telegram n8n node sends the result or confirmation back to the user.

To deploy this powerful n8n workflow template, follow these steps:

1. Import: Copy the provided JSON and import it directly into your n8n instance using the 'New' button and 'Import from JSON'.


2. Credentials Setup: This n8n workflow requires credentials for three services:

1. Database Configuration: In the Notion n8n node steps (Create a database page, Get Notion Tasks (analyze), Get Notion Tasks (update)), replace the placeholder Database ID with the ID of your Notion task database.


2. Activate Trigger: The Receive Telegram Messages n8n trigger must be configured with the correct webhook URL, which n8n will generate upon activation. Once configured, set the entire n8n workflow to 'Active'.

This n8n template relies on sophisticated LangChain and Notion n8n node configurations:

Receive Telegram Messages (n8n Trigger): The starting n8n node, listening for incoming messages (text or voice) to the configured Telegram bot.
Voice or Text? (Switch n8n node): Routes execution based on the presence of $json.message.voice.file_id or $json.message.text.
Transcribe Voice to Text (OpenAI n8n node): Uses OpenAI's audio resource with the translate operation to convert spoken word into text, a crucial step for handling voice inputs in this n8n workflow.
Detect Intent (ChainLLM n8n node): A classification model using gpt-4o-mini to determine if the user intends to create, update, or analyze tasks. This decision powers the conditional flow of the n8n workflow.
Create Task (Prompt) (ChainLLM n8n node + Output Parser): Extracts structured data (title, date, priority) from unstructured text input. It is paired with an Output Parser n8n node to ensure the output is valid JSON for Notion.
Get Notion Tasks (Notion n8n node): Fetches all existing task pages from the specified Notion database, used both for the update path (for semantic matching) and the analyze path (for generating LLM insights).
Find Matching Task (ChainLLM n8n node): Uses the search string and the full task list to precisely identify the Notion item the user wants to modify.
Update a database page (HTTP Request n8n node): Performs a complex PATCH request to the Notion API, dynamically constructing the properties object (Name, Description, Priority, Due Date, Tags) based on the AI-generated update JSON. This ensures the n8n automation handles complex data types correctly.

• AI Analyze Tasks (ChainLLM n8n node): Generates human-like, reflective analysis and advice based on the user's question and their task list, demonstrating the high value of LLMs in this n8n workflow.