This is an n8n workflow that automates outbound customer calls with AI agents. The agents hold two-way conversations in English, Hindi, and Gujarati. It uses Twilio for calling, Anthropic Claude for language processing, Supabase for appointment data, and ElevenLabs for voice.
The workflow runs automatically every hour and reads appointment details from Supabase. It builds a call context message with customer details, service time, and appointment status, then starts an outbound call through Twilio. Claude writes replies in the caller's language.
Based on what the caller says, the workflow can confirm, reschedule, or cancel an appointment through Supabase MCP. ElevenLabs text-to-speech delivers the voice replies through TwiML webhooks. It is intended for appointment confirmations, reminders, and follow-ups, such as in customer support, healthcare and service scheduling, and SaaS operations.
- Languages
- English, Hindi, Gujarati
- Schedule
- Runs every hour
- Data source
- Supabase (appointment details)
- Telephony
- Twilio outbound calls
- AI model
- Anthropic Claude
- Voice
- ElevenLabs text-to-speech via TwiML webhooks
- Appointment actions
- Confirm, reschedule, cancel (via Supabase MCP)
- n8n nodes
- Set, Code, Webhook, Sticky Note, HTTP Request, n8n-nodes-langchain agent
- Credentials needed
- Twilio, Anthropic, Supabase API, Supabase MCP, ElevenLabs