← All N8N workflows
AI & researchAgentChain LLMDocument Default Data LoaderEmbeddings Open AiFilter

AI Phone Agent with RetellAI

Configured with Agent, Chain LLM, Document Default Data Loader, Embeddings Open Ai, and additional nodes. Review the connections and setup before adapting this template.

Workflow preview
RoutingFilterFilterAI & modelsOpenAI Chat ModelOpenAI Chat ModelTransformStructured OutputParserOutput Parser StructuredTriggern8n_rag_functionWebhookData storeRetrive Qdrant VectorStoreVector Store QdrantAI & modelsEmbeddings OpenAI2Embeddings Open AiAI & modelsRAGTool Vector StoreAI & modelsOpenAI Chat Model2OpenAI Chat ModelWeb / APIRespond to WebhookRespond To WebhookAI & modelsOpenAI Chat Model1OpenAI Chat ModelMessagingTelegramTelegramIntegrationGoogle CalendarGoogle CalendarAI & modelsOpenAI Chat Model3OpenAI Chat ModelTransformStructured OutputParser1Output Parser StructuredTriggern8n_callWebhookTriggern8n_check_availableWebhookTriggerWhen clicking ‘Testworkflow’Manual TriggerData storeQdrant Vector StoreVector Store QdrantWeb / APICreate collectionHTTP RequestWeb / APIRefresh collectionHTTP RequestFiles / mediaGet folderGoogle DriveFiles / mediaDownload FilesGoogle DriveAI & modelsEmbeddings OpenAIEmbeddings Open AiFiles / mediaDefault Data LoaderDocument Default Data Lo…TransformToken SplitterText Splitter Token Spli…TransformSet call fieldsEdit FieldsAI & modelsExtract key pointsChain LLMAI & modelsConcert start dateChain LLMAI & modelsRetrive AgentAgent

Node categories in this workflow

  • Routing
  • AI & models
  • Transform
  • Trigger
  • Data store
  • Web / API
  • Messaging
  • Integration
  • Files / media

Icons describe general node roles, inferred from node types. Inspect the parameters for the configured action.

Drag to pan; use + and − to zoom. Reconstructed from the exported node positions; nothing runs here. Dashed lines represent non-main connections. Full node names and notes appear below.

How it works

The exported workflow contains the following configured nodes. Their connections are shown above; inspect the JSON for the complete logic, expressions, and external actions.

Workflow node names, types, and operations
NodeCategoryTypeOperation
FilterRoutingFilterSee node parameters
OpenAI Chat ModelAI & modelsOpenAI Chat ModelSee node parameters
Structured Output ParserTransformOutput Parser StructuredSee node parameters
n8n_rag_functionTriggerWebhookSee node parameters
Retrive Qdrant Vector StoreData storeVector Store QdrantSee node parameters
Embeddings OpenAI2AI & modelsEmbeddings Open AiSee node parameters
RAGAI & modelsTool Vector StoreSee node parameters
OpenAI Chat Model2AI & modelsOpenAI Chat ModelSee node parameters
Respond to WebhookWeb / APIRespond To WebhookSee node parameters
OpenAI Chat Model1AI & modelsOpenAI Chat ModelSee node parameters
TelegramMessagingTelegramSee node parameters
Google CalendarIntegrationGoogle CalendarSee node parameters
OpenAI Chat Model3AI & modelsOpenAI Chat ModelSee node parameters
Structured Output Parser1TransformOutput Parser StructuredSee node parameters
n8n_callTriggerWebhookSee node parameters
n8n_check_availableTriggerWebhookSee node parameters
When clicking ‘Test workflow’TriggerManual TriggerSee node parameters
Qdrant Vector StoreData storeVector Store QdrantSee node parameters
Create collectionWeb / APIHTTP RequestPOST
Refresh collectionWeb / APIHTTP RequestPOST
Get folderFiles / mediaGoogle DrivefileFolder
Download FilesFiles / mediaGoogle Drivedownload
Embeddings OpenAIAI & modelsEmbeddings Open AiSee node parameters
Default Data LoaderFiles / mediaDocument Default Data LoaderSee node parameters
Token SplitterTransformText Splitter Token SplitterSee node parameters
Set call fieldsTransformEdit FieldsSee node parameters
Extract key pointsAI & modelsChain LLMSee node parameters
Concert start dateAI & modelsChain LLMSee node parameters
Retrive AgentAI & modelsAgentSee node parameters

Setup

  1. Download the JSON and import it into an isolated n8n instance. Leave the workflow inactive.
  2. Connect your own credentials through n8n’s credential store. Replace REPLACE_WITH_* values with your resource identifiers.
  3. Inspect API destinations, Code nodes, database queries, file paths, and outbound actions.
  4. Run with synthetic data and harmless destinations. Enable triggers only after checking the actual outputs.

Requirements

An n8n instance with the listed node types and compatible versions. Service-connected workflows may require accounts, credentials, or paid APIs.

Credential types declared in the source

  • googleCalendarOAuth2Api
  • googleDriveOAuth2Api
  • httpHeaderAuth
  • openAiApi
  • qdrantApi
  • telegramApi

Workflow notes

STEP 3 - RETELL AI

STEP 3 - RETELL AI

  • Register on Retell AI (10$ FREE credits)
  • Create an Agent an set "Voice & Language" and add your system prompt
  • In Webhook settings add the "Agent Level Webhook URL" with the n8n webhook node url called "n8n_call"
  • Buy a new phone number with your FREE credits by Twilio Provider and connect it to the created agent
  • Enter the previously created agency and create the flow as shown in the following image
  • Aggiungere 2 funzioni (una per RAG e una per il Booking) e inserire l'url apposito ricavato dai webhook di n8n "n8n_rag_function" e "n8n_check_available"
STEP 4

STEP 4

Intercept the "end call" event and get the full call transcript

  • Add your CHAT_ID in Telegram node
STEP 1

STEP 1

Create Qdrant Collection

Change:

  • QDRANTURL
  • COLLECTION
STEP 2

STEP 2

Documents vectorization with Qdrant and Google Drive

Change:

  • QDRANTURL
  • COLLECTION
STEP 5

STEP 5

If required, create the event in the calendar

  • Enter the title and description of the event
STEP 6

STEP 6

If required retrive the informations by RAG system

Create your first AI Phone Agent

Create your first AI Phone Agent

Build, test, deploy, and monitor AI phone agents. Retell is a comprehensive platform for building, testing, deploying, and monitoring reliable AI phone agents. Conversation flow agent allows you to create multiple nodes to handle different scenarios in the conversation. It provides more fine-grained control over the conversation flow compared to single / multi prompt agent, which unlocks the ability to handle more complex scenarios.

This Workflow simulates an AI-powered phone agent with two main functions:

📅 Appointment Booking – It can schedule appointments directly into Google Calendar.

🧠 RAG-based Information Retrieval – It provides answers using a Retrieval-Augmented Generation (RAG) system. For example, it can respond to questions such as store opening hours, return policies, or product details.

The guide also explains how to purchase a dedicated phone number (with a +1 prefix) and link it to the AI agent. This setup is cost-effective, as it uses a free $10 credit to operate without additional charges in the beginning.

AI Phone Agent with RetellAI

AI Phone Agent with RetellAI

Purpose and structure

This folder contains the AI Phone Agent with RetellAI workflow. Its executable nodes use agent, chainLlm, documentDefaultDataLoader, embeddingsOpenAi, filter, googleCalendar, googleDrive, httpRequest, lmChatOpenAi, manualTrigger, outputParserStructured, respondToWebhook, set, telegram, textSplitterTokenSplitter, toolVectorStore, vectorStoreQdrant, webhook. The table and connection map below are derived from the JSON, not an inferred guarantee of behavior.

Setup

Import inactive. Configure your own credentials and replace REPLACE_WITH_* values. Review external destinations and Code nodes before running.

Entry points

n8n_rag_function, n8n_call, n8n_check_available, When clicking ‘Test workflow’

Verification

Sanitized source candidate; not runtime-tested. This overview is generated from node metadata. Read the folder README for the connection map, credential types, and license.

Attribution

https://github.com/Zie619/n8n-workflows/blob/94007c1445d9258a7da116646b79473e7c7c3282/workflows/Telegram/0735_Telegram_GoogleCalendar_Automation_Webhook.json MIT; preserve source notices.

More in AI & research