{
  "nodes": [
    {
      "parameters": {},
      "id": "a6149859-d2e7-4b16-9ef7-46f9e486e4d3",
      "name": "When clicking Test workflow",
      "type": "n8n-nodes-base.manualTrigger",
      "typeVersion": 1,
      "position": [
        64,
        -80
      ]
    },
    {
      "parameters": {
        "resource": "fileFolder",
        "returnAll": true,
        "filter": {
          "folderId": {
            "__rl": true,
            "value": "REPLACE_WITH_RESOURCE",
            "mode": "id"
          }
        },
        "options": {}
      },
      "id": "ca005ba5-6ff4-4ea5-acda-3af378733dea",
      "name": "Google Drive - List Files1",
      "type": "n8n-nodes-base.googleDrive",
      "typeVersion": 3,
      "position": [
        288,
        -80
      ]
    },
    {
      "parameters": {
        "operation": "executeQuery",
        "query": "delete from documents where source_file = $1;",
        "options": {
          "queryBatching": "independently",
          "queryReplacement": "={{ [$json.name] }}"
        }
      },
      "type": "n8n-nodes-base.postgres",
      "typeVersion": 2.6,
      "position": [
        512,
        -80
      ],
      "id": "31fecc35-a37a-4e1c-8b5e-80a4ad10060e",
      "name": "Delete old chunks",
      "executeOnce": false
    },
    {
      "parameters": {
        "operation": "download",
        "fileId": {
          "__rl": true,
          "value": "={{ $('Google Drive - List Files1').all()[$itemIndex].json.id }}",
          "mode": "id"
        },
        "options": {}
      },
      "id": "126ea62c-bc07-4740-bdbf-118ed4d7b473",
      "name": "Google Drive - Download File1",
      "type": "n8n-nodes-base.googleDrive",
      "typeVersion": 3,
      "position": [
        736,
        -80
      ]
    },
    {
      "parameters": {
        "operation": "pdf",
        "binaryPropertyName": "data",
        "options": {}
      },
      "id": "f3a8c2d1-9e4b-4c6a-8f21-3d7b5a9c1e02",
      "name": "Extract PDF Text",
      "type": "n8n-nodes-base.extractFromFile",
      "typeVersion": 1,
      "position": [
        960,
        -80
      ]
    },
    {
      "parameters": {
        "jsCode": "// The upstream 'Extract PDF Text' node outputs the real extracted text\n// in item.json.text. We look up the original filename by item index from\n// the List Files node (rather than relying on binary metadata surviving\n// the extraction step) so department-tagging still works reliably.\nconst output = [];\nconst items = $input.all();\n\nfor (let i = 0; i < items.length; i++) {\n  const item = items[i];\n  const fileName = $('Google Drive - List Files1').all()[i].json.name || 'unknown.txt';\n  const content = item.json.text || '';\n\n  let department = 'general';\n  if (fileName.startsWith('billing_')) department = 'billing';\n  else if (fileName.startsWith('engineering_')) department = 'engineering';\n  else if (fileName.startsWith('support_')) department = 'support';\n\n  output.push({ json: { source_file: fileName, department, content } });\n}\n\nreturn output;"
      },
      "id": "87ea7b6c-a217-4ea2-bc7d-76b4337c91ee",
      "name": "Parse File to Text1",
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [
        1184,
        -80
      ]
    },
    {
      "parameters": {
        "jsCode": "// Chunk each document into ~200-word pieces with 40% overlap\nconst CHUNK_SIZE = 200;\nconst OVERLAP = 80;\n\nconst output = [];\n\nfor (const item of $input.all()) {\n  const { source_file, department, content } = item.json;\n  const words = content.split(/\\s+/);\n\n  let start = 0;\n  let chunkIndex = 0;\n\n  while (start < words.length) {\n    const end = Math.min(start + CHUNK_SIZE, words.length);\n    const chunkText = words.slice(start, end).join(' ');\n\n    output.push({\n      json: {\n        source_file,\n        department,\n        chunk_index: chunkIndex,\n        content: chunkText\n      }\n    });\n\n    chunkIndex++;\n    if (end === words.length) break;\n    start = end - OVERLAP;\n  }\n}\n\nreturn output;"
      },
      "id": "37e2c771-5f96-43fc-94d2-0fb7ae2dbe19",
      "name": "Chunk Text1",
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [
        1408,
        -80
      ]
    },
    {
      "parameters": {
        "options": {}
      },
      "type": "n8n-nodes-base.splitInBatches",
      "typeVersion": 3,
      "position": [
        1632,
        -80
      ],
      "id": "5e3a754f-d710-411e-b511-ffd01f831d7e",
      "name": "Loop Over Items"
    },
    {
      "parameters": {
        "method": "POST",
        "url": "http://host.docker.internal:11434/api/embeddings",
        "sendBody": true,
        "specifyBody": "json",
        "jsonBody": "={\n  \"model\": \"nomic-embed-text\",\n  \"prompt\": {{ JSON.stringify($json.content) }}\n}",
        "options": {}
      },
      "id": "71128970-b4ae-47ca-ad90-67cefd03d57a",
      "name": "Ollama - Embed Chunk1",
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 4.2,
      "position": [
        1856,
        -176
      ]
    },
    {
      "parameters": {
        "jsCode": "// Merge the embedding result back with the original chunk metadata\nconst originalItems = $('Chunk Text1').all();\nconst embedItems = $input.all();\n\nconst output = [];\n\nfor (let i = 0; i < embedItems.length; i++) {\n  const original = originalItems[i].json;\n  const embedding = embedItems[i].json.embedding.values;\n\n  output.push({\n    json: {\n      content: original.content,\n      source_file: original.source_file,\n      department: original.department,\n      chunk_index: original.chunk_index,\n      embedding: embedding\n    }\n  });\n}\n\nreturn output;"
      },
      "id": "bac81d61-d385-42f0-ab83-05e78db20fc4",
      "name": "Merge Embedding + Metadata1",
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [
        1856,
        16
      ]
    },
    {
      "parameters": {
        "operation": "executeQuery",
        "query": "insert into documents (content, embedding, source_file, department, chunk_index)\nvalues ($1, $2::vector, $3, $4, $5);",
        "options": {
          "queryReplacement": "={{ [$json.content, JSON.stringify($json.embedding), $json.source_file, $json.department, $json.chunk_index] }}"
        }
      },
      "id": "30b6d21f-74dd-4422-bfd0-51456ae20f51",
      "name": "Supabase - Replace Chunk1",
      "type": "n8n-nodes-base.postgres",
      "typeVersion": 2.5,
      "position": [
        2080,
        16
      ]
    },
    {
      "id": "c42b9640-ad30-51b8-b533-d06b6ce66a3b",
      "name": "Publication · Overview and setup",
      "type": "n8n-nodes-base.stickyNote",
      "typeVersion": 1,
      "position": [
        -800,
        -176
      ],
      "parameters": {
        "content": "# google drive list files1 and executequery postgres and download google drive\n\n## Purpose and structure\nThis folder contains the **google drive list files1 and executequery postgres and download google drive** workflow. Its executable nodes use `code`, `extractFromFile`, `googleDrive`, `httpRequest`, `manualTrigger`, `postgres`, `splitInBatches`. The table and connection map below are derived from the JSON, not an inferred guarantee of behavior.\n\n## Setup\nImport inactive. Configure your own credentials and replace REPLACE_WITH_* values. Review external destinations and Code nodes before running.\n\n## Entry points\nWhen clicking Test workflow\n\n## Verification\nSanitized source candidate; not runtime-tested. This overview is generated from node metadata. Read the folder README for the connection map, credential types, and license.\n\n## Attribution\nhttps://github.com/TahsinTanni/RAG-slackbot/blob/f51615aee91eb029228fc0a9853df0ad2cd5d0b0/index.json\nMIT; preserve source notices.",
        "width": 700,
        "height": 800,
        "color": 1
      }
    }
  ],
  "connections": {
    "When clicking Test workflow": {
      "main": [
        [
          {
            "node": "Google Drive - List Files1",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Google Drive - List Files1": {
      "main": [
        [
          {
            "node": "Delete old chunks",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Delete old chunks": {
      "main": [
        [
          {
            "node": "Google Drive - Download File1",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Google Drive - Download File1": {
      "main": [
        [
          {
            "node": "Extract PDF Text",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Extract PDF Text": {
      "main": [
        [
          {
            "node": "Parse File to Text1",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Parse File to Text1": {
      "main": [
        [
          {
            "node": "Chunk Text1",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Chunk Text1": {
      "main": [
        [
          {
            "node": "Loop Over Items",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Loop Over Items": {
      "main": [
        [
          {
            "node": "Merge Embedding + Metadata1",
            "type": "main",
            "index": 0
          }
        ],
        [
          {
            "node": "Ollama - Embed Chunk1",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Ollama - Embed Chunk1": {
      "main": [
        [
          {
            "node": "Loop Over Items",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Merge Embedding + Metadata1": {
      "main": [
        [
          {
            "node": "Supabase - Replace Chunk1",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  },
  "active": false
}