🎧 IT Voice Support Automation Bot – Telegram Voice Message to JIRA ticket with OpenAI Whisper

Automatically process IT support requests submitted via Telegram voice messages by transcribing, extracting structured data, creating a JIRA ticket, and notifying relevant parties.

🧑‍💼 Who’s it for

• Internal teams that handle IT support but want to streamline voice-based requests.
• Employees who prefer using mobile/voice to report incidents or ask for support.
• Organizations aiming to integrate conversational AI into existing support workflows.

⚙️ How it works / What it does

1. A user sends a voice message to a Telegram bot.
2. The system checks whether it’s an audio message.
3. If valid, the audio is:
   • Downloaded
   • Transcribed via OpenAI Whisper
   • Backed up to Google Drive
4. The transcription and file metadata are merged.
5. The merged content is processed through an AI Agent (GPT) to extract structured request info.
6. A JIRA ticket is created using the extracted data.
7. The IT team is notified via Slack (or other channels).
8. The requester receives a Telegram confirmation message with the JIRA ticket link.
9. If the input is not audio, a polite rejection message is sent.

📌 Key Features

• Supports voice-based ticket creation
• Accurate transcription using Whisper
• Context-aware request parsing using GPT-4.1 mini
• Fully automated ticket creation in JIRA
• Notifies both IT and the original requester
• Cloud backup of original voice messages (Google Drive)

🛠️ Setup Instructions

Prerequisites

Workflow Steps

1. Telegram Voice Message Trigger: Starts the flow when a user sends a voice message.

2. Is Audio Message?: If false → reply "only voice is supported"

3. Download Audio: Download .oga file from Telegram.

4. Transcribe Audio: Use OpenAI Whisper to get text transcript.

5. Backup to Google Drive: Upload original voice file with metadata.

6. Merge Results: Combine transcript and metadata.

7. Pre-process Output: Clean formatting before AI extraction.

8. Transcript Processing Agent: GPT-based agent extracts:

   • Requester name, department
   • Request title & description
   • Priority & request type

9. Submit JIRA Request Ticket: Create ticket from AI-extracted data.

10. Setup Slack / Email / Manual Steps: Optional internal routing or approvals.

11. Inform Reporter via Telegram: Sends confirmation message with JIRA ticket link.

🔧 How to Customize

• Replace JIRA with Zendesk, GitHub Issues, or other ticketing tools.
• Change Slack to Microsoft Teams or Email.
• Add Notion/Airtable logging.
• Enhance agent to extract department from user ID or metadata.

📦 Requirements

Built with ❤️ using n8n