{ "edges": [ { "to": "91bd04ed-60f2-448e-8031-4a33616f78a0", "from": "bd24942e-1d59-4a95-9f7f-ac9e64835737", "toPort": "0", "fromPort": "ai_languageModel" }, { "to": "dcc02c60-9600-4bcd-a772-655ce3d937b5", "from": "7d6011fd-40c0-491b-ba28-4a51147ac23f", "toPort": "0", "fromPort": "0" }, { "to": "fa7eb92a-48f6-4b06-b270-aeb2510e12cd", "from": "dcc02c60-9600-4bcd-a772-655ce3d937b5", "toPort": "0", "fromPort": "0" }, { "to": "ea186a1a-6d32-458b-b715-911a6fd551c0", "from": "91bd04ed-60f2-448e-8031-4a33616f78a0", "toPort": "0", "fromPort": "0" }, { "to": "91bd04ed-60f2-448e-8031-4a33616f78a0", "from": "fa7eb92a-48f6-4b06-b270-aeb2510e12cd", "toPort": "0", "fromPort": "0" }, { "to": "305d9aa9-e043-412a-a8ef-ec5c807960ba", "from": "ea186a1a-6d32-458b-b715-911a6fd551c0", "toPort": "0", "fromPort": "0" }, { "to": "46137cfe-cb9a-4851-911b-d2baae8e840b", "from": "ea186a1a-6d32-458b-b715-911a6fd551c0", "toPort": "0", "fromPort": "0" }, { "to": "f3aa5cef-9c71-483a-9ed0-0d5f5bd0c00f", "from": "305d9aa9-e043-412a-a8ef-ec5c807960ba", "toPort": "0", "fromPort": "0" } ], "nodes": [ { "id": "7d6011fd-40c0-491b-ba28-4a51147ac23f", "name": "Daily feedback check", "type": "trigger", "params": { "kind": "n8n-nodes-base.scheduleTrigger", "rule": { "interval": [ { "triggerAtHour": 9 } ] } }, "position": [ 128, 0 ] }, { "id": "dcc02c60-9600-4bcd-a772-655ce3d937b5", "name": "Read feedback responses", "type": "integration", "params": { "service": "googleSheets", "operation": "read", "sheetName": { "__rl": true, "mode": "list", "value": "" }, "documentId": { "__rl": true, "mode": "list", "value": "" } }, "position": [ 304, 0 ] }, { "id": "fa7eb92a-48f6-4b06-b270-aeb2510e12cd", "name": "Filter unprocessed responses", "type": "code", "params": { "jsCode": "// Filter unprocessed rows and batch them for AI analysis\nvar items = $input.all();\nvar unprocessed = [];\nfor (var i = 0; i < items.length; i++) {\n var row = items[i].json;\n if (!row.Processed || row.Processed === '' || row.Processed === 'no') {\n unprocessed.push({\n rowIndex: i + 2,\n timestamp: row.Timestamp || '',\n response: row.Feedback || row.Response || row['Your feedback'] || '',\n rating: row.Rating || row.Score || ''\n });\n }\n}\nif (unprocessed.length === 0) {\n return { json: { skip: true, message: 'No new responses to analyze' } };\n}\nvar feedbackText = unprocessed.map(function(r, idx) {\n return 'Response ' + (idx + 1) + ' (row ' + r.rowIndex + '): ' + r.response + (r.rating ? ' [Rating: ' + r.rating + ']' : '');\n}).join('\\n');\nreturn { json: { skip: false, count: unprocessed.length, feedbackText: feedbackText, rows: unprocessed } };" }, "position": [ 464, 0 ] }, { "id": "91bd04ed-60f2-448e-8031-4a33616f78a0", "name": "Analyze feedback with Gemini", "type": "llm", "params": { "text": "=Analyze the following customer feedback responses. For each response, determine:\n1. Sentiment: positive, neutral, or negative\n2. Sentiment score: 1-10 (1=very negative, 10=very positive)\n3. Category: product, service, pricing, usability, support, or other\n4. Key theme: a short phrase summarizing the main point\n\nAlso provide an overall summary with:\n- Total responses analyzed\n- Sentiment distribution (count of positive/neutral/negative)\n- Top 3 recurring themes\n- One actionable recommendation\n\nReturn valid JSON with this structure:\n{\n \"responses\": [{\"rowIndex\": number, \"sentiment\": string, \"score\": number, \"category\": string, \"theme\": string}],\n \"summary\": {\"total\": number, \"positive\": number, \"neutral\": number, \"negative\": number, \"topThemes\": [string], \"recommendation\": string}\n}\n\nFeedback data:\n{{ $json.feedbackText }}", "prompt": "=Analyze the following customer feedback responses. For each response, determine:\n1. Sentiment: positive, neutral, or negative\n2. Sentiment score: 1-10 (1=very negative, 10=very positive)\n3. Category: product, service, pricing, usability, support, or other\n4. Key theme: a short phrase summarizing the main point\n\nAlso provide an overall summary with:\n- Total responses analyzed\n- Sentiment distribution (count of positive/neutral/negative)\n- Top 3 recurring themes\n- One actionable recommendation\n\nReturn valid JSON with this structure:\n{\n \"responses\": [{\"rowIndex\": number, \"sentiment\": string, \"score\": number, \"category\": string, \"theme\": string}],\n \"summary\": {\"total\": number, \"positive\": number, \"neutral\": number, \"negative\": number, \"topThemes\": [string], \"recommendation\": string}\n}\n\nFeedback data:\n{{ $json.feedbackText }}", "provider": "openai", "promptType": "define", "approximated": true }, "position": [ 672, 0 ] }, { "id": "bd24942e-1d59-4a95-9f7f-ac9e64835737", "name": "Gemini Chat Model", "type": "llm", "params": { "options": { "temperature": 0.3 }, "provider": "google", "modelName": "models/gemini-2.0-flash", "temperature": 0.3 }, "position": [ 672, 144 ] }, { "id": "ea186a1a-6d32-458b-b715-911a6fd551c0", "name": "Parse results and build report", "type": "code", "params": { "jsCode": "// Parse AI response and prepare data for Sheets and email\nvar aiResponse = $input.first().json;\nvar text = aiResponse.text || aiResponse.response || JSON.stringify(aiResponse);\nvar parsed;\ntry {\n var jsonMatch = text.match(/\\{[\\s\\S]*\\}/);\n parsed = JSON.parse(jsonMatch ? jsonMatch[0] : text);\n} catch(e) {\n parsed = { responses: [], summary: { total: 0, positive: 0, neutral: 0, negative: 0, topThemes: [], recommendation: 'Could not parse AI response' } };\n}\nvar responses = parsed.responses || [];\nvar summary = parsed.summary || {};\nvar rows = [];\nfor (var i = 0; i < responses.length; i++) {\n var r = responses[i];\n rows.push({\n rowIndex: r.rowIndex || 0,\n sentiment: r.sentiment || 'unknown',\n score: r.score || 5,\n category: r.category || 'other',\n theme: r.theme || ''\n });\n}\nvar positiveRate = summary.total > 0 ? Math.round((summary.positive / summary.total) * 100) : 0;\nvar negativeRate = summary.total > 0 ? Math.round((summary.negative / summary.total) * 100) : 0;\nvar emailBody = '

Customer Feedback Analysis Report

';\nemailBody += '

Date: ' + new Date().toISOString().split('T')[0] + '

';\nemailBody += '

Responses analyzed: ' + (summary.total || 0) + '

';\nemailBody += '

Sentiment Distribution

';\nemailBody += '

Positive: ' + (summary.positive || 0) + ' (' + positiveRate + '%)
Neutral: ' + (summary.neutral || 0) + '
Negative: ' + (summary.negative || 0) + ' (' + negativeRate + '%)

';\nemailBody += '

Top Themes

' + themes[t] + '

';\nemailBody += '

Recommendation

' + (summary.recommendation || 'N/A') + '

';\nreturn { json: { rows: rows, summary: summary, positiveRate: positiveRate, negativeRate: negativeRate, emailBody: emailBody, topThemes: (themes).join(', '), recommendation: summary.recommendation || 'N/A', total: summary.total || 0 } };" }, "position": [ 1120, 0 ] }, { "id": "305d9aa9-e043-412a-a8ef-ec5c807960ba", "name": "Update feedback sheet with results", "type": "integration", "params": { "service": "googleSheets", "operation": "update", "sheetName": { "__rl": true, "mode": "list", "value": "" }, "documentId": { "__rl": true, "mode": "list", "value": "" } }, "position": [ 1344, 0 ] }, { "id": "46137cfe-cb9a-4851-911b-d2baae8e840b", "name": "Log insights to Notion", "type": "integration", "params": { "options": {}, "service": "notion", "resource": "databasePage", "operation": "read", "databaseId": { "__rl": true, "mode": "list", "value": "" }, "propertiesUi": { "propertyValues": [ { "key": "Date" }, { "key": "Total Responses", "numberValue": "={{ $json.total }}" }, { "key": "Positive Rate", "numberValue": "={{ $json.positiveRate }}" }, { "key": "Top Themes" }, { "key": "Recommendation" } ] } }, "position": [ 1568, 0 ] }, { "id": "f3aa5cef-9c71-483a-9ed0-0d5f5bd0c00f", "name": "Email report to team", "type": "integration", "params": { "message": "={{ $json.emailBody }}", "options": {}, "service": "gmail", "subject": "=Customer Feedback Report - {{ new Date().toISOString().split('T')[0] }}", "operation": "read" }, "position": [ 1568, 224 ] } ] }

Analyze Google Forms feedback with Gemini AI and send weekly reports