Voice-Activated Gmail Response System using OpenAI Whisper and Telegram - n8n Workflow

• Professionals managing high email volumes who need rapid response generation.


• Users who prefer dictating responses rather than typing.


• Technical users looking for practical, real-world examples of complex n8n templates combining AI, Gmail, and Telegram.


• Developers interested in building custom AI-powered assistants using the n8n node ecosystem.

Managing a busy inbox often requires quick, thoughtful replies, but typing them out can be time-consuming. This sophisticated n8n workflow solves this by allowing users to respond to critical emails using only their voice via Telegram.

The system works in two major stages. First, the Gmail n8n trigger monitors your inbox, and an OpenAI agent intelligently filters out spam or automated emails, ensuring only those requiring a genuine response proceed. Second, when the user sends a voice note reply via Telegram, the n8n node sequence processes the audio using OpenAI Whisper for high-fidelity transcription. Finally, a second large language model generates a polished, context-aware email draft linked to the original thread in Gmail. This seamless integration showcases the power of a customized n8n automation for improving personal productivity.

This automation operates using two distinct n8n trigger points and multiple AI stages:

1. Email Intake (Gmail Trigger): The n8n workflow starts when the Gmail n8n trigger detects a new email. An initial 'If' n8n node ensures only messages in the primary INBOX are processed.


2. AI Classification: A LangChain LLM chain, powered by the OpenAI Chat Model, determines if the email needs a reply (Y/N) based on predefined criteria (e.g., filtering out newsletters and 'no-reply' senders).


3. Telegram Alert: If a reply is needed, the email metadata (Subject, Sender, Thread ID) is sent to a specified user or group via the Telegram n8n node.


4. Voice Response Trigger: The second part of the n8n workflow is activated by the Telegram Bot Message Received n8n trigger. An 'If' n8n node validates that the incoming Telegram message is both a reply to the email notification and contains an audio file (voice note).


5. Transcription (OpenAI Whisper): The 'Get Audio File' n8n node retrieves the voice note, which is then sent to the OpenAI n8n node (using the Whisper API operation) for accurate transcription.


6. Draft Generation: The transcribed text and the original email context are fed into the 'Write Polished Reply' LangChain LLM chain. This chain uses an OpenAI model to transform the casual voice note into a professional email response.


7. Draft Creation (Gmail): Finally, the 'Create Email Draft' n8n node uses the original thread ID to create a new, polished draft reply directly in Gmail, ready for final review and sending.


8. Confirmation: A link to the newly created Gmail draft is sent back to the user via Telegram, completing the n8n workflow cycle.

To deploy this specialized n8n workflow, follow these steps:

1. Import the n8n templates: Download the provided JSON and import it into your n8n instance using the 'Workflows' > 'New' > 'Import from JSON' option.


2. Configure Credentials: You will need to set up credentials for the following services:

1. Set Telegram Chat ID: Locate the 'Set Chat ID' n8n node. You must replace the placeholder value with your specific Telegram Chat ID. This is the recipient where email alerts will be sent.


2. Activate the n8n trigger nodes: Ensure both the 'New Email Received' (Gmail Trigger) and 'Telegram Bot Message Received' (Telegram Trigger) n8n nodes are active and configured correctly.


3. Save and Activate: Save the complete n8n workflow and set it to 'Active'.

New Email Received (Gmail Trigger): The initial n8n trigger. It polls the Gmail API for new messages, acting as the starting point for the email processing chain.
Does Email Need a Response? (LangChain LLM Chain): This powerful n8n node uses an internal prompt and the Open AI Chat Model to classify the incoming email, directing the n8n workflow flow based on necessity.
Structured Output Parser (LangChain): Attached to the LLM chain, this ensures the AI output is strictly a usable JSON object ({ response: "Y" } or N), which allows the subsequent 'Needs a response?' If n8n node to function reliably.
Text Email (Telegram n8n node): Responsible for formatting the original email data—including crucial thread and recipient IDs—and transmitting it to the designated user chat.
Telegram Bot Message Received (Telegram Trigger): The second n8n trigger, initiating the voice-to-draft sequence when a new message is received by the bot.
OpenAI (LangChain OpenAI n8n node): Configured for the 'audio' resource and 'transcribe' operation. This n8n node leverages the robust OpenAI Whisper API to convert the user's voice note into accurate text.
Write Polished Reply (LangChain LLM Chain): Takes the transcribed voice note and the original email context (pulled dynamically from the Telegram reply) to generate a professionally worded email draft.
Create Email Draft (Gmail n8n node): A critical action n8n node that uses the generated text to create a draft in Gmail, ensuring it is correctly linked to the corresponding original email thread via the Thread ID parameter.