David Olusola This workflow corresponds to n8n.io template #12568 — we link there as the canonical source.
This workflow follows the Agent → Gmail Tool 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 →
{
"meta": {
"templateCredsSetupCompleted": true
},
"nodes": [
{
"id": "3b741f12-c482-4cad-add1-d982b4bae2d9",
"name": "\ud83d\udccc START HERE",
"type": "n8n-nodes-base.stickyNote",
"position": [
-1264,
1024
],
"parameters": {
"color": 4,
"width": 516,
"height": 664,
"content": "## \ud83e\udd16 Telegram assistant with memory\n\n**Author:** **David Olusola (@DaexAI)**\n\nAI assistant that remembers conversations across sessions using MemMachine.\n\n**Features:**\n\u2705 Persistent cross-session memory\n\u2705 Voice transcription (Whisper)\n\u2705 Gmail, Sheets, Calendar tools\n\u2705 MCP architecture\n\n## How it works\n1. Message arrives (text/voice)\n2. Store in MemMachine\n3. Retrieve conversation history\n4. AI processes with context\n5. Store response & send\n\n## Setup steps\n1. Install MemMachine: `docker-compose up -d`\n2. Update org_id/project_id (nodes 4,5,9)\n3. Add credentials (Telegram, OpenAI, Google)\n4. Test with messages"
},
"typeVersion": 1
},
{
"id": "9b2a7af6-54ea-42bb-8eab-8022659b293b",
"name": "Voice Processing",
"type": "n8n-nodes-base.stickyNote",
"position": [
-640,
1440
],
"parameters": {
"color": 7,
"width": 540,
"height": 180,
"content": "**Voice path:** Download \u2192 Transcribe (Whisper) \u2192 Extract data"
},
"typeVersion": 1
},
{
"id": "dd33c807-8d40-4e4d-a895-79e78c7d6cc4",
"name": "Text Processing",
"type": "n8n-nodes-base.stickyNote",
"position": [
-176,
1680
],
"parameters": {
"color": 7,
"height": 140,
"content": "**Text path:** Extract message data"
},
"typeVersion": 1
},
{
"id": "5cfcad20-0260-43e2-b2c0-14a5869f1138",
"name": "MemMachine Memory",
"type": "n8n-nodes-base.stickyNote",
"position": [
32,
1536
],
"parameters": {
"color": 5,
"width": 560,
"height": 200,
"content": "**Memory system:** Store message \u2192 Search history (30 memories) \u2192 Sort chronologically"
},
"typeVersion": 1
},
{
"id": "673d48ec-9108-4fe1-8cb1-1ba4ff9114db",
"name": "AI Processing",
"type": "n8n-nodes-base.stickyNote",
"position": [
720,
1488
],
"parameters": {
"color": 6,
"width": 280,
"height": 388,
"content": "**AI + Tools:** Process with context, access Gmail/Sheets/Calendar via MCP"
},
"typeVersion": 1
},
{
"id": "ffd7e526-7280-412e-81e5-540b820613a7",
"name": "Response Flow",
"type": "n8n-nodes-base.stickyNote",
"position": [
1232,
1536
],
"parameters": {
"color": 3,
"width": 716,
"height": 232,
"content": "**Response:** Extract \u2192 Prepare \u2192 Store in memory \u2192 Send to Telegram"
},
"typeVersion": 1
},
{
"id": "4f2d4409-42ac-43b9-a18f-8560e5e808d3",
"name": "MCP Architecture",
"type": "n8n-nodes-base.stickyNote",
"position": [
1120,
992
],
"parameters": {
"color": 2,
"width": 480,
"height": 392,
"content": "**MCP Server:** Exposes Gmail, Sheets, Calendar. **MCP Client:** Connects AI to tools"
},
"typeVersion": 1
},
{
"id": "e162b482-4c08-4f7e-a804-c6266e846218",
"name": "Memory Example",
"type": "n8n-nodes-base.stickyNote",
"position": [
-704,
1024
],
"parameters": {
"color": 2,
"width": 420,
"height": 400,
"content": "## \ud83c\udfaf Memory Example\n\n**Day 1:** \"Email john@co.com about Q1\"\n**Day 3:** \"What did I ask about John?\"\n**AI:** \"You asked me to email John about Q1 on Jan 5th\"\n\n**Day 7 (voice):** \"Follow up with John\"\n**AI:** \"I'll send follow-up to john@co.com about Q1\"\n\nAI remembers who, what, when across text & voice!"
},
"typeVersion": 1
},
{
"id": "d20a1ce4-8ff8-4602-add3-f4349d7199e4",
"name": "Quick Setup",
"type": "n8n-nodes-base.stickyNote",
"position": [
-240,
1024
],
"parameters": {
"color": 5,
"width": 564,
"height": 480,
"content": "## \ud83d\ude80 Quick Setup\n\n**1. Install MemMachine**\n```bash\ngit clone [https://github.com/MemMachine/MemMachine](https://github.com/MemMachine/MemMachine)\ncd MemMachine\ndocker-compose up -d\n```\n\n**2. Import & Configure**\n- Import this workflow\n- Update nodes 4,5,9: org_id, project_id\n- Add credentials (Telegram, OpenAI, Google)\n\n**3. Test**\n- Text message \u2192 Check response\n- Voice message \u2192 Check transcription\n- Ask about past chats \u2192 Check memory"
},
"typeVersion": 1
},
{
"id": "4394aac4-e55c-4d13-be11-975ce5573825",
"name": "Customization",
"type": "n8n-nodes-base.stickyNote",
"position": [
448,
1024
],
"parameters": {
"color": 6,
"width": 360,
"height": 320,
"content": "## \ud83d\udca1 Customization\n\n**Memory:** Increase top_k (node 5)\n**Tools:** Add Notion, Slack, Trello\n**Multi-channel:** WhatsApp, SMS, Web\n**AI:** Edit system prompt (node 7)\n**Advanced:** User preferences, automation"
},
"typeVersion": 1
},
{
"id": "5cfd7a21-60e3-49de-8a27-cd2ca26cdb58",
"name": "1. Telegram Trigger",
"type": "n8n-nodes-base.telegramTrigger",
"position": [
-1040,
1616
],
"parameters": {
"updates": [
"message"
],
"additionalFields": {}
},
"typeVersion": 1.2
},
{
"id": "d97b3cb4-9a24-4020-9019-91fc19616ce9",
"name": "2. Message Type",
"type": "n8n-nodes-base.switch",
"position": [
-816,
1616
],
"parameters": {
"rules": {
"values": [
{
"outputKey": "Voice",
"conditions": {
"options": {
"version": 2
},
"conditions": [
{
"operator": {
"type": "string",
"operation": "exists"
},
"leftValue": "={{ $json.message.voice.file_id }}"
}
]
},
"renameOutput": true
},
{
"outputKey": "Text",
"conditions": {
"options": {
"version": 2
},
"conditions": [
{
"operator": {
"type": "string",
"operation": "exists"
},
"leftValue": "={{ $json.message.text }}"
}
]
},
"renameOutput": true
}
]
},
"options": {}
},
"typeVersion": 3.2
},
{
"id": "1eb5a7b4-92e9-41e1-951d-0f6cf2ffb053",
"name": "3a. Download Voice",
"type": "n8n-nodes-base.telegram",
"position": [
-592,
1520
],
"parameters": {
"fileId": "={{ $json.message.voice.file_id }}",
"resource": "file",
"additionalFields": {}
},
"typeVersion": 1.2
},
{
"id": "371ce3d2-18c8-481b-a9be-f92b789a4743",
"name": "3b. Transcribe Voice",
"type": "@n8n/n8n-nodes-langchain.openAi",
"position": [
-368,
1520
],
"parameters": {
"options": {},
"resource": "audio",
"operation": "transcribe"
},
"typeVersion": 1.6
},
{
"id": "29e37fd3-95fc-4339-8d06-3dba2f8b924c",
"name": "3c. Extract Voice Data",
"type": "n8n-nodes-base.set",
"position": [
-144,
1520
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "a1",
"name": "text",
"type": "string",
"value": "={{ $json.text }}"
},
{
"id": "a2",
"name": "chat_id",
"type": "number",
"value": "={{ $('1. Telegram Trigger').item.json.message.chat.id }}"
},
{
"id": "a3",
"name": "user_id",
"type": "number",
"value": "={{ $('1. Telegram Trigger').item.json.message.from.id }}"
},
{
"id": "a4",
"name": "username",
"type": "string",
"value": "={{ $('1. Telegram Trigger').item.json.message.from.username || $('1. Telegram Trigger').item.json.message.from.first_name }}"
},
{
"id": "a5",
"name": "customer_email",
"type": "string",
"value": "=dimejicole21@gmail.com"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "b86ed0d2-be45-432f-a22a-740bf3a67fd7",
"name": "3d. Extract Text Data",
"type": "n8n-nodes-base.set",
"position": [
-144,
1712
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "a1",
"name": "text",
"type": "string",
"value": "={{ $json.message.text }}"
},
{
"id": "a2",
"name": "chat_id",
"type": "number",
"value": "={{ $json.message.chat.id }}"
},
{
"id": "a3",
"name": "user_id",
"type": "number",
"value": "={{ $json.message.from.id }}"
},
{
"id": "a4",
"name": "username",
"type": "string",
"value": "={{ $json.message.from.username || $json.message.from.first_name }}"
},
{
"id": "a5",
"name": "customer_email",
"type": "string",
"value": "=dimejicole21@gmail.com"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "8485263c-d446-479e-8d74-480574dea53b",
"name": "Unifier",
"type": "n8n-nodes-base.code",
"position": [
64,
1616
],
"parameters": {
"jsCode": "return $input.all();"
},
"typeVersion": 2
},
{
"id": "c189a91d-c337-4fe4-a3eb-d9c776b67f21",
"name": "4. Store User Query",
"type": "n8n-nodes-base.httpRequest",
"position": [
240,
1616
],
"parameters": {
"url": "http://host.docker.internal:8080/api/v2/memories",
"method": "POST",
"options": {},
"jsonBody": "={\n \"org_id\": \"insuracrm\",\n \"project_id\": \"crm-chat-v6\",\n \"messages\": [{\n \"content\": {{ JSON.stringify($json.text) }},\n \"producer\": {{ JSON.stringify($json.customer_email) }},\n \"produced_for\": \"assistant\",\n \"role\": \"user\",\n \"metadata\": {\n \"customer_email\": {{ JSON.stringify($json.customer_email) }},\n \"channel\": \"telegram\",\n \"username\": {{ JSON.stringify($json.username) }},\n \"user_id\": {{ JSON.stringify(String($json.user_id)) }},\n \"timestamp\": {{ JSON.stringify($now.toISO()) }}\n }\n }]\n}",
"sendBody": true,
"specifyBody": "json"
},
"typeVersion": 4.1
},
{
"id": "bd92e6bb-8794-49d9-8a76-076ebe2e68ee",
"name": "5. Search Memory",
"type": "n8n-nodes-base.httpRequest",
"position": [
432,
1616
],
"parameters": {
"url": "http://host.docker.internal:8080/api/v2/memories/search",
"method": "POST",
"options": {},
"jsonBody": "={\n \"org_id\": \"insuracrm\",\n \"project_id\": \"crm-chat-v6\",\n \"query\": \"{{ $('Unifier').item.json.text }}\",\n \"top_k\": 30,\n \"filter\": \"metadata.customer_email='{{ $('Unifier').item.json.customer_email }}'\",\n \"types\": [\"episodic\", \"semantic\"]\n}",
"sendBody": true,
"specifyBody": "json"
},
"typeVersion": 4.1
},
{
"id": "8a7166f2-82c4-437e-aaed-0bf9a77955c7",
"name": "6. Format Memory",
"type": "n8n-nodes-base.code",
"position": [
608,
1616
],
"parameters": {
"jsCode": "const memoryData = $input.first().json;\nconst episodicMemory = memoryData.content?.episodic_memory || {};\nconst longTermEpisodes = episodicMemory.long_term_memory?.episodes || [];\nconst shortTermEpisodes = episodicMemory.short_term_memory?.episodes || [];\nconst allMemories = [...longTermEpisodes, ...shortTermEpisodes];\n\nconst sortedMemories = allMemories.sort((a, b) => {\n const timeA = a.created_at || a.metadata?.timestamp || '';\n const timeB = b.created_at || b.metadata?.timestamp || '';\n return new Date(timeA).getTime() - new Date(timeB).getTime();\n});\n\nconst conversationHistory = sortedMemories\n .filter(m => m.content && m.content.trim() !== '')\n .slice(-15)\n .map(memory => {\n const role = memory.producer_role || 'unknown';\n const channel = memory.metadata?.channel || 'telegram';\n return `[${role} via ${channel}]: ${memory.content}`;\n }).join('\\n');\n\nconst userData = $('Unifier').first().json;\n\nreturn [{ json: { ...userData, conversation_history: conversationHistory || 'No history.', total_memories: allMemories.length } }];"
},
"typeVersion": 2
},
{
"id": "96f443ba-e716-4d52-bc2b-b0b4b62a57b1",
"name": "7. AI Agent",
"type": "@n8n/n8n-nodes-langchain.agent",
"position": [
800,
1616
],
"parameters": {
"text": "={{ $json.text }}",
"options": {
"systemMessage": "=# Role: InsuraOS Intelligence Core\nYou are the primary assistant for InsuraOS.\n\nUser: {{ $json.username }}\nCustomer: {{ $json.customer_email }}\n\n## \ud83e\udde0 Memory Context\n{{ $json.conversation_history }}\n\n## Instructions\n1. Maintain a single-room conversation feel.\n2. NEVER mention technical channel tags in the reply.\n\nTime: {{ $now }}"
},
"promptType": "define"
},
"typeVersion": 1.7
},
{
"id": "a26d6d5b-9ecc-46a2-8bc4-8ca3944dba03",
"name": "8. Extract Response",
"type": "n8n-nodes-base.set",
"position": [
1056,
1616
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "a1",
"name": "ai_response",
"type": "string",
"value": "={{ $json.output }}"
},
{
"id": "a2",
"name": "customer_email",
"type": "string",
"value": "={{ $('6. Format Memory').item.json.customer_email }}"
},
{
"id": "a3",
"name": "username",
"type": "string",
"value": "={{ $('6. Format Memory').item.json.username }}"
},
{
"id": "a4",
"name": "chat_id",
"type": "number",
"value": "={{ $('6. Format Memory').item.json.chat_id }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "774e5725-9817-4589-827d-133e3549eaec",
"name": "8b. Prepare for Storage",
"type": "n8n-nodes-base.code",
"position": [
1200,
1616
],
"parameters": {
"jsCode": "const inputData = $input.first().json;\nconst payload = {\n org_id: \"insuracrm\",\n project_id: \"crm-chat-v6\",\n messages: [{\n content: inputData.ai_response || '',\n producer: \"assistant\",\n produced_for: inputData.customer_email || '',\n role: \"assistant\",\n metadata: {\n customer_email: inputData.customer_email || '',\n channel: \"telegram\",\n username: inputData.username || '',\n timestamp: new Date().toISOString()\n }\n }]\n};\nreturn [{ json: { payload, ai_response: inputData.ai_response, chat_id: inputData.chat_id } }];"
},
"typeVersion": 2
},
{
"id": "09cd456b-dcbb-41fd-b45d-96314e467397",
"name": "9. Store AI Response",
"type": "n8n-nodes-base.httpRequest",
"position": [
1376,
1616
],
"parameters": {
"url": "http://host.docker.internal:8080/api/v2/memories",
"method": "POST",
"options": {},
"jsonBody": "={{ JSON.stringify($json.payload) }}",
"sendBody": true,
"specifyBody": "json"
},
"typeVersion": 4.1
},
{
"id": "f2f482d5-e89b-4cae-a095-fc5d1db8fdcc",
"name": "10. Send Response",
"type": "n8n-nodes-base.telegram",
"position": [
1568,
1616
],
"parameters": {
"text": "={{ $json.ai_response }}",
"chatId": "={{ $json.chat_id }}",
"additionalFields": {
"parse_mode": "Markdown",
"appendAttribution": false
}
},
"typeVersion": 1.2
},
{
"id": "23bddf97-05b1-4db7-89f8-afa93f7c6d0b",
"name": "MCP Server Trigger",
"type": "@n8n/n8n-nodes-langchain.mcpTrigger",
"position": [
1264,
1040
],
"parameters": {
"path": "298e8ee6-95f2-408b-97df-af61f48a20ec"
},
"typeVersion": 2
},
{
"id": "5489706c-c1a5-4320-b045-2f57bb6fde4e",
"name": "Gmail Tool",
"type": "n8n-nodes-base.gmailTool",
"position": [
1200,
1248
],
"parameters": {
"options": {}
},
"typeVersion": 2.2
},
{
"id": "4cf74443-3727-4b9c-8ac8-94c2f3d4f4f3",
"name": "Google Sheets Tool",
"type": "n8n-nodes-base.googleSheetsTool",
"position": [
1328,
1248
],
"parameters": {
"sheetName": {
"__rl": true,
"mode": "list",
"value": ""
},
"documentId": {
"__rl": true,
"mode": "list",
"value": ""
}
},
"typeVersion": 4.7
},
{
"id": "f4c07b00-1243-4dbe-a54b-e052edbdb9d5",
"name": "Google Calendar Tool",
"type": "n8n-nodes-base.googleCalendarTool",
"position": [
1456,
1232
],
"parameters": {
"calendar": {
"__rl": true,
"mode": "list",
"value": ""
},
"additionalFields": {}
},
"typeVersion": 1.3
},
{
"id": "346e67fa-e6c0-4cd0-9ecc-cdf7df73d56f",
"name": "MCP Client",
"type": "@n8n/n8n-nodes-langchain.mcpClientTool",
"position": [
896,
1792
],
"parameters": {
"options": {}
},
"typeVersion": 1.2
},
{
"id": "a537219d-c0bb-4dda-949b-27987e8ae5d6",
"name": "OpenAI Chat Model",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"position": [
768,
1792
],
"parameters": {
"model": "gpt-4o-mini",
"options": {},
"builtInTools": {}
},
"typeVersion": 1.3
}
],
"connections": {
"Unifier": {
"main": [
[
{
"node": "4. Store User Query",
"type": "main",
"index": 0
}
]
]
},
"Gmail Tool": {
"ai_tool": [
[
{
"node": "MCP Server Trigger",
"type": "ai_tool",
"index": 0
}
]
]
},
"MCP Client": {
"ai_tool": [
[
{
"node": "7. AI Agent",
"type": "ai_tool",
"index": 0
}
]
]
},
"7. AI Agent": {
"main": [
[
{
"node": "8. Extract Response",
"type": "main",
"index": 0
}
]
]
},
"2. Message Type": {
"main": [
[
{
"node": "3a. Download Voice",
"type": "main",
"index": 0
}
],
[
{
"node": "3d. Extract Text Data",
"type": "main",
"index": 0
}
]
]
},
"5. Search Memory": {
"main": [
[
{
"node": "6. Format Memory",
"type": "main",
"index": 0
}
]
]
},
"6. Format Memory": {
"main": [
[
{
"node": "7. AI Agent",
"type": "main",
"index": 0
}
]
]
},
"OpenAI Chat Model": {
"ai_languageModel": [
[
{
"node": "7. AI Agent",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"3a. Download Voice": {
"main": [
[
{
"node": "3b. Transcribe Voice",
"type": "main",
"index": 0
}
]
]
},
"Google Sheets Tool": {
"ai_tool": [
[
{
"node": "MCP Server Trigger",
"type": "ai_tool",
"index": 0
}
]
]
},
"1. Telegram Trigger": {
"main": [
[
{
"node": "2. Message Type",
"type": "main",
"index": 0
}
]
]
},
"4. Store User Query": {
"main": [
[
{
"node": "5. Search Memory",
"type": "main",
"index": 0
}
]
]
},
"8. Extract Response": {
"main": [
[
{
"node": "8b. Prepare for Storage",
"type": "main",
"index": 0
}
]
]
},
"3b. Transcribe Voice": {
"main": [
[
{
"node": "3c. Extract Voice Data",
"type": "main",
"index": 0
}
]
]
},
"9. Store AI Response": {
"main": [
[
{
"node": "10. Send Response",
"type": "main",
"index": 0
}
]
]
},
"Google Calendar Tool": {
"ai_tool": [
[
{
"node": "MCP Server Trigger",
"type": "ai_tool",
"index": 0
}
]
]
},
"3d. Extract Text Data": {
"main": [
[
{
"node": "Unifier",
"type": "main",
"index": 0
}
]
]
},
"3c. Extract Voice Data": {
"main": [
[
{
"node": "Unifier",
"type": "main",
"index": 0
}
]
]
},
"8b. Prepare for Storage": {
"main": [
[
{
"node": "9. Store AI Response",
"type": "main",
"index": 0
}
]
]
}
}
}
Pro
For the full experience including quality scoring and batch install features for each workflow upgrade to Pro
About this workflow
An AI assistant that NEVER forgets using MemMachine for persistent cross-session memory, with voice transcription support and productivity tools.
Source: https://n8n.io/workflows/12568/ — 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.
Generate AI viral videos with NanoBanana & VEO3, shared on socials via Blotato 2. Uses @blotato/n8n-nodes-blotato, googleSheets, lmChatOpenAi, toolThink. Event-driven trigger; 94 nodes.
@Blotato/N8N Nodes Blotato, Google Sheets, OpenAI Chat +9
Telegram Trigger receives incoming messages (text, voice, photo, document). Switch routes by message type to appropriate processors: Text → forwarded as-is. Voice → downloaded and sent to Transcribe a
Memory Buffer Window, Telegram Trigger, Telegram +12
Transform your Telegram messenger into a powerful, multi-modal personal or team assistant. This n8n workflow creates an intelligent agent that can understand text, voice, images, and documents, and ta
Memory Buffer Window, Telegram Trigger, Telegram +10
Digital marketers, content creators, social media managers, and businesses who want to use AI marketing automation for YouTube Shorts without spending hours on production. This AI workflow helps anyon
OpenAI, HTTP Request, OpenAI Chat +7
Transform your salon/service business with this streamlined Telegram automation system featuring Claude integration, zero-setup database management, and intelligent conversation handling. Claude MCP I
Redis, Agent Tool, Google Calendar +10