This workflow follows the Agent → OpenAI Embeddings recipe pattern — see all workflows that pair these two integrations.
The workflow JSON
Copy or download the full n8n JSON below. Paste it into a new n8n workflow, add your credentials, activate. Full import guide →
{
"name": "IA - Indexa\u00e7\u00e3o Google Docs -> Supabase pgvector",
"nodes": [
{
"parameters": {
"content": "## Fluxo de atendimento (runtime)\n\nRoda a cada mensagem recebida no WhatsApp.\n\n1. Recebe texto ou \u00e1udio\n2. Transcreve \u00e1udio (se houver)\n3. Busca contexto no Supabase (pgvector)\n4. Agente responde com o contexto encontrado, ou avisa que n\u00e3o sabe\n5. Envia resposta pelo WhatsApp\n\nO conte\u00fado do Google Docs \u00e9 indexado por um workflow separado, agendado \u2014 veja \"n8n-indexacao-google-docs.json\".",
"height": 404,
"width": 380
},
"id": "d8b07f03-6a12-4796-9beb-36e638250193",
"name": "Nota - Como funciona",
"type": "n8n-nodes-base.stickyNote",
"typeVersion": 1,
"position": [
-1008,
384
]
},
{
"parameters": {
"updates": [
"messages"
],
"options": {}
},
"id": "4f132f57-27e6-4850-82cd-09d682b003d3",
"name": "1. WhatsApp Trigger",
"type": "n8n-nodes-base.whatsAppTrigger",
"typeVersion": 1,
"position": [
-448,
672
]
},
{
"parameters": {
"conditions": {
"options": {
"caseSensitive": true,
"leftValue": "",
"typeValidation": "strict",
"version": 2
},
"conditions": [
{
"id": "cond-audio",
"leftValue": "={{ $json.messages[0].type }}",
"rightValue": "audio",
"operator": {
"type": "string",
"operation": "equals"
}
}
],
"combinator": "and"
},
"options": {}
},
"id": "919d2151-ad36-4b05-abda-27d4ac87d46f",
"name": "2. Tem \u00e1udio?",
"type": "n8n-nodes-base.if",
"typeVersion": 2.2,
"position": [
-144,
672
]
},
{
"parameters": {
"url": "=https://graph.facebook.com/v19.0/{{ $json.messages[0].audio.id }}",
"authentication": "predefinedCredentialType",
"nodeCredentialType": "whatsAppApi",
"options": {
"response": {
"response": {
"responseFormat": "file"
}
}
}
},
"id": "20243a56-fc93-4203-bc07-2ffc5b1e7895",
"name": "3a. Baixar \u00e1udio",
"type": "n8n-nodes-base.httpRequest",
"typeVersion": 4.2,
"position": [
-32,
400
]
},
{
"parameters": {
"assignments": {
"assignments": [
{
"id": "set-pergunta-texto",
"name": "pergunta",
"type": "string",
"value": "={{ $json.messages[0].text.body }}"
},
{
"id": "set-telefone",
"name": "telefone",
"type": "string",
"value": "={{ $json.messages[0].from }}"
}
]
},
"options": {}
},
"id": "4b8cfea5-a198-4b76-8b56-da51c2971369",
"name": "3c. Extrair texto",
"type": "n8n-nodes-base.set",
"typeVersion": 3.4,
"position": [
304,
688
]
},
{
"parameters": {
"assignments": {
"assignments": [
{
"id": "set-pergunta-audio",
"name": "pergunta",
"type": "string",
"value": "={{ $json.text }}"
},
{
"id": "set-telefone-audio",
"name": "telefone",
"type": "string",
"value": "={{ $('1. WhatsApp Trigger').item.json.messages[0].from }}"
}
]
},
"options": {}
},
"id": "a5a825fa-93ae-4bfb-8a62-b427c3de9c3c",
"name": "3d. Normalizar campo pergunta",
"type": "n8n-nodes-base.set",
"typeVersion": 3.4,
"position": [
528,
688
]
},
{
"parameters": {
"model": {
"value": "text-embedding-3-small"
},
"options": {}
},
"id": "41a15893-d2f4-4bb6-b931-8c98d1f0fe61",
"name": "Embeddings para busca",
"type": "@n8n/n8n-nodes-langchain.embeddingsOpenAi",
"typeVersion": 1.2,
"position": [
688,
912
]
},
{
"parameters": {
"mode": "load",
"tableName": "documents",
"prompt": "={{ $json.pergunta }}",
"options": {}
},
"id": "75f6fa9b-52b5-451a-a8ba-1c9060a9ad20",
"name": "4. Buscar no Supabase",
"type": "@n8n/n8n-nodes-langchain.vectorStoreSupabase",
"typeVersion": 1.3,
"position": [
704,
688
],
"credentials": {
"supabaseApi": {
"name": "<your credential>"
}
}
},
{
"parameters": {
"jsCode": "// Avalia os resultados da busca vetorial e monta o contexto para o agente.\n// Ajuste o LIMIAR conforme os testes com sua base (0.75 \u00e9 um bom ponto de partida\n// para embeddings text-embedding-3-small com dist\u00e2ncia de cosseno).\nconst LIMIAR_SIMILARIDADE = 0.75;\n\nconst resultados = items.map(item => item.json);\n\nconst relevantes = resultados.filter(r => (r.score ?? r.similarity ?? 0) >= LIMIAR_SIMILARIDADE);\n\nconst contexto = relevantes\n .map(r => r.pageContent || r.content || '')\n .join('\\n\\n---\\n\\n');\n\nreturn [{\n json: {\n pergunta: $('4. Buscar no Supabase').item.json.pergunta || items[0]?.json?.pergunta,\n telefone: $input.first().json.telefone,\n contexto,\n encontrouContexto: relevantes.length > 0\n }\n}];"
},
"id": "c78b35d9-17b4-4089-a1e8-4718c87231e5",
"name": "5. Avaliar resultados",
"type": "n8n-nodes-base.code",
"typeVersion": 2,
"position": [
1024,
688
]
},
{
"parameters": {
"model": {
"value": "gpt-4o-mini"
},
"options": {}
},
"id": "c843ee32-d32f-47a7-b00c-f56e9c4e4906",
"name": "Modelo de LLM",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"typeVersion": 1.2,
"position": [
1248,
928
]
},
{
"parameters": {
"promptType": "define",
"text": "={{ $json.pergunta }}",
"options": {
"systemMessage": "=Voc\u00ea \u00e9 um assistente de suporte que responde perguntas do usu\u00e1rio usando SOMENTE o contexto abaixo, extra\u00eddo da base de conhecimento da empresa.\n\nContexto encontrado (pode estar vazio):\n{{ $json.contexto }}\n\nRegras:\n- Se o contexto responder \u00e0 pergunta, responda de forma clara e completa usando essas informa\u00e7\u00f5es.\n- Se o contexto estiver vazio ou n\u00e3o tiver rela\u00e7\u00e3o com a pergunta, diga honestamente que n\u00e3o encontrou essa informa\u00e7\u00e3o na base de conhecimento e sugira que o usu\u00e1rio reformule a pergunta ou fale com um atendente humano.\n- Nunca invente informa\u00e7\u00f5es que n\u00e3o estejam no contexto."
}
},
"id": "209aa905-f922-464f-ae27-4909e3058e80",
"name": "6. Agente de IA",
"type": "@n8n/n8n-nodes-langchain.agent",
"typeVersion": 1.9,
"position": [
1248,
688
]
},
{
"parameters": {
"operation": "send",
"phoneNumberId": "SUBSTITUIR-PHONE-NUMBER-ID",
"recipientPhoneNumber": "={{ $('5. Avaliar resultados').item.json.telefone }}",
"textBody": "={{ $json.output }}",
"additionalFields": {}
},
"id": "7a4c0b05-e5df-4a10-b589-229d83d542dc",
"name": "7. Enviar resposta",
"type": "n8n-nodes-base.whatsApp",
"typeVersion": 1,
"position": [
1632,
688
]
},
{
"parameters": {
"resource": "audio",
"operation": "transcribe",
"options": {}
},
"type": "@n8n/n8n-nodes-langchain.openAi",
"typeVersion": 2.3,
"position": [
192,
400
],
"id": "c1a46b31-c926-431d-bf67-091105b2fd3d",
"name": "Transcribe a recording"
},
{
"parameters": {},
"type": "@n8n/n8n-nodes-langchain.memoryPostgresChat",
"typeVersion": 1.4,
"position": [
1344,
912
],
"id": "d741afa6-81a9-4f91-98af-9c5d442794d0",
"name": "Postgres Chat Memory"
},
{
"parameters": {
"operation": "get"
},
"type": "n8n-nodes-base.googleDocsTool",
"typeVersion": 2,
"position": [
1536,
928
],
"id": "516bc5e5-2506-4ba6-b4ff-fd5d69e70178",
"name": "Get a document in Google Docs",
"credentials": {
"googleDocsOAuth2Api": {
"name": "<your credential>"
}
}
}
],
"connections": {
"1. WhatsApp Trigger": {
"main": [
[
{
"node": "2. Tem \u00e1udio?",
"type": "main",
"index": 0
}
]
]
},
"2. Tem \u00e1udio?": {
"main": [
[
{
"node": "3a. Baixar \u00e1udio",
"type": "main",
"index": 0
}
],
[
{
"node": "3c. Extrair texto",
"type": "main",
"index": 0
}
]
]
},
"3c. Extrair texto": {
"main": [
[
{
"node": "3d. Normalizar campo pergunta",
"type": "main",
"index": 0
}
]
]
},
"3d. Normalizar campo pergunta": {
"main": [
[
{
"node": "4. Buscar no Supabase",
"type": "main",
"index": 0
}
]
]
},
"Embeddings para busca": {
"ai_embedding": [
[
{
"node": "4. Buscar no Supabase",
"type": "ai_embedding",
"index": 0
}
]
]
},
"4. Buscar no Supabase": {
"main": [
[
{
"node": "5. Avaliar resultados",
"type": "main",
"index": 0
}
]
]
},
"5. Avaliar resultados": {
"main": [
[
{
"node": "6. Agente de IA",
"type": "main",
"index": 0
}
]
]
},
"Modelo de LLM": {
"ai_languageModel": [
[
{
"node": "6. Agente de IA",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"6. Agente de IA": {
"main": [
[
{
"node": "7. Enviar resposta",
"type": "main",
"index": 0
}
]
]
},
"Transcribe a recording": {
"main": [
[
{
"node": "3c. Extrair texto",
"type": "main",
"index": 0
}
]
]
},
"3a. Baixar \u00e1udio": {
"main": [
[
{
"node": "Transcribe a recording",
"type": "main",
"index": 0
}
]
]
},
"Postgres Chat Memory": {
"ai_memory": [
[
{
"node": "6. Agente de IA",
"type": "ai_memory",
"index": 0
}
]
]
},
"Get a document in Google Docs": {
"ai_tool": [
[
{
"node": "6. Agente de IA",
"type": "ai_tool",
"index": 0
}
]
]
}
},
"active": false,
"settings": {
"executionOrder": "v1",
"binaryMode": "separate"
},
"versionId": "d8e6579e-8214-43e4-92bf-b83975263230",
"nodeGroups": [],
"id": "ip9iFA8mms6q9kxK",
"tags": []
}
Credentials you'll need
Each integration node will prompt for credentials when you import. We strip credential IDs before publishing — you'll add your own.
googleDocsOAuth2ApisupabaseApi
Pro
For the full experience including quality scoring and batch install features for each workflow upgrade to Pro
About this workflow
IA - Indexação Google Docs -> Supabase pgvector. Uses whatsAppTrigger, httpRequest, embeddingsOpenAi, vectorStoreSupabase. Event-driven trigger; 15 nodes.
Source: https://github.com/CodeX984/PORTIFOLIO/blob/0af9890ad5b581134a5b53625754ab337e3de78f/Agente-Suporte/workflows/Fluxo-Atendimento.json — original creator credit. Request a take-down →
Related workflows
Workflows that share integrations, category, or trigger type with this one. All free to copy and import.
📌 Overview
Redis, WhatsApp, OpenAI Chat +12
Your AI workforce is ready. Are you?
Google Sheets Tool, Mcp Trigger, Google Drive +29
This intelligent chatbot leverages cutting-edge financial APIs and AI-driven analysis to deliver comprehensive stock research reports. Get instant access to professional-grade investment analysis that
Tool Think, Supabase Vector Store, OpenAI Embeddings +15
___MAIN_2_Modularizado_24_3___ATIVO___ADVOCACIA_n8n___AGENTE_ENCAMINHAMENTO_. Uses supabase, httpRequest, informationExtractor, lmChatOpenAi. Event-driven trigger; 68 nodes.
Supabase, HTTP Request, Information Extractor +11
Who is this for? This workflow is ideal for HR teams, startups, and enterprises that want to handle employee interactions through WhatsApp and automate responses using LLM (OpenAI) and intelligent rou
WhatsApp Trigger, OpenAI, OpenAI Chat +13