{ "id": "x1D6kcIROn2c236i", "meta": { "templateCredsSetupCompleted": true }, "name": "CV Rating Automation - Indeed", "tags": [], "nodes": [ { "id": "ff78627a-9d2e-4439-8f6a-00b488f7cff5", "name": "\ud83d\udccb Overview Workflow", "type": "n8n-nodes-base.stickyNote", "position": [ 400, 640 ], "parameters": { "color": 3, "width": 380, "height": 756, "content": "## \ud83c\udfaf CV Rating Automation - Production\n\n**Full automated candidate screening pipeline for Indeed applications.**\n\n### Pipeline\n1. IMAP receives Indeed application email\n2. Metadata extraction (job position, candidate name)\n3. CV upload to Google Drive (folder per position)\n4. OCR with Mistral AI (base64 inline)\n5. Text cleaning & normalization\n6. LLM analysis with Groq + Llama 3.3 70B + Structured Output Parser\n7. Score \u2194 recommendation consistency validation\n8. Routing by score:\n - \u226575 \u2192 detailed email to recruiter\n - always \u2192 append to Google Sheets\n9. Error branch \u2192 ops notification\n\n### Required Stack\n- n8n + PostgreSQL + Redis (Docker)\n- Queue mode with workers for parallelism\n\n### Required Credentials\n- IMAP (Indeed mailbox)\n- Google Drive OAuth2\n- Google Sheets OAuth2\n- Mistral AI API\n- Groq API\n- SMTP (for notifications)\n\n### Estimated Cost\n~\u20ac0.005 per processed CV (OCR + LLM)" }, "typeVersion": 1 }, { "id": "6a569f42-5944-4096-88b8-8a2cb1011bf7", "name": "\u2699\ufe0f Configuration", "type": "n8n-nodes-base.set", "position": [ 976, 1008 ], "parameters": { "options": {}, "assignments": { "assignments": [ { "id": "config-drive-folder", "name": "DRIVE_ROOT_FOLDER_ID", "type": "string", "value": "YOUR_GOOGLE_DRIVE_FOLDER_ID" }, { "id": "config-sheet-id", "name": "SHEET_ID", "type": "string", "value": "YOUR_GOOGLE_SHEET_ID" }, { "id": "config-recruiter-email", "name": "RECRUITER_EMAIL", "type": "string", "value": "user@example.com" }, { "id": "config-ops-email", "name": "OPS_EMAIL", "type": "string", "value": "user@example.com" }, { "id": "config-sender-email", "name": "SENDER_EMAIL", "type": "string", "value": "user@example.com" }, { "id": "config-score-threshold", "name": "SCORE_THRESHOLD", "type": "number", "value": 75 }, { "id": "config-imap-subject-filter", "name": "IMAP_SUBJECT_FILTER", "type": "string", "value": "New application for" } ] } }, "typeVersion": 3.4 }, { "id": "4864c888-d142-4baf-8454-74192720eb6b", "name": "\ud83d\udce5 Note IMAP", "type": "n8n-nodes-base.stickyNote", "position": [ 800, 576 ], "parameters": { "color": 5, "width": 336, "height": 616, "content": "## \ud83d\udce5 IMAP Trigger\n\n**Why IMAP instead of a webhook?**\nIndeed does not offer public APIs to receive applications in real time. A dedicated mailbox is the most reliable solution.\n\n### Recommended Setup\n- Dedicated mailbox (e.g. `applications@company.com`)\n- Filter `subjectContains: application`\n- Polling every 1 minute\n- `downloadAttachments: true`\n\n### Alternatives\n- Indeed Apply API (enterprise employers)\n- ATS bridge (Greenhouse, Workable)\n- Dedicated email parser (Mailparser)" }, "typeVersion": 1 }, { "id": "6a6877b9-593b-4792-86e4-15cfb92b8590", "name": "\ud83d\udd0d Note Extract", "type": "n8n-nodes-base.stickyNote", "position": [ 1152, 640 ], "parameters": { "color": 5, "width": 336, "height": 712, "content": "## \ud83d\udd0d Metadata Extraction\n\nThe Code node extracts:\n- **Job position** (regex on Indeed subject line)\n- **Candidate name** (trailing part of the subject)\n- **Valid attachments** (PDF/DOCX only)\n\n\u26a0\ufe0f **Adapt the regex** to match the email format you receive: the Indeed template varies by language. Test with real emails.\n\n\ud83d\udca1 Emails without valid attachments are **silently dropped**." }, "typeVersion": 1 }, { "id": "a14f0f3d-3631-4907-bacf-6d967bf60a8a", "name": "\ud83d\udcc1 Note Drive", "type": "n8n-nodes-base.stickyNote", "position": [ 1504, 512 ], "parameters": { "color": 5, "width": 336, "height": 752, "content": "## \ud83d\udcc1 Google Drive Organization\n\nCVs are saved in **per-position folders**:\n```\nCV_Senior_Backend_Developer/\nCV_Frontend_React/\n```\n\n\u26a0\ufe0f **Replace `YOUR_FOLDER_ID`** with the ID of your root folder on Drive (found in the folder URL).\n\n### Benefits\n- Organized historical archive\n- Easy sharing with HR team\n- Direct link in Google Sheets\n- GDPR compliance (easy bulk deletion)" }, "typeVersion": 1 }, { "id": "1b2f75bb-21e3-4921-958b-39822f34ef56", "name": "\ud83e\udde0 Note OCR", "type": "n8n-nodes-base.stickyNote", "position": [ 1872, 432 ], "parameters": { "color": 4, "width": 368, "height": 820, "content": "## \ud83e\udde0 OCR with Mistral\n\nUses **`mistral-ocr-latest`**, optimized for document processing.\n\n### Strategy: inline base64\nThe PDF is sent to Mistral **as base64 in the request body** instead of via a Drive URL. Benefits:\n- \u2705 No dependency on public Drive permissions\n- \u2705 More secure (no public files)\n- \u2705 Faster (zero round-trip)\n- \u2705 50MB limit is more than enough for CVs\n\n### Cost\n~\u20ac0.001 per page, ~\u20ac0.003 per average CV.\n\n### Optional Optimization\nFor simple text-only CVs (no graphics), add a branch with `Extract from File` (free) before Mistral." }, "typeVersion": 1 }, { "id": "8fba51de-d1e8-472e-8426-e7900d5fb1b0", "name": "\u2728 Note AI Chain", "type": "n8n-nodes-base.stickyNote", "position": [ 2272, 544 ], "parameters": { "color": 4, "width": 452, "height": 860, "content": "## \u2728 Core AI Flow\n\nCluster of 3 LangChain nodes:\n```\nCV Analyzer Chain (root)\n \u251c\u2500 Groq Chat Model \u2190 ai_languageModel\n \u2514\u2500 Output Parser \u2190 ai_outputParser\n```\n\n### Benefits\n- **Groq**: ultra-fast inference (10x faster than OpenAI)\n- **Structured Output Parser**: injects JSON Schema into the prompt and auto-retries on non-compliant output\n- **Always valid JSON output** according to schema\n\n\ud83d\udca1 To switch models (`gpt-oss-120b`, `mixtral-8x7b`), simply change the dropdown in the Groq node." }, "typeVersion": 1 }, { "id": "b95b4eaa-0794-4cad-9fe1-73536adcd90e", "name": "\ud83d\udcd0 Note Schema", "type": "n8n-nodes-base.stickyNote", "position": [ 2752, 496 ], "parameters": { "color": 4, "width": 320, "height": 668, "content": "## \ud83d\udcd0 Output Schema\n\nThe JSON Schema defines 16 typed fields:\n\n**Personal Info**: first_name, last_name, email, phone\n\n**Profile**: years_experience, education, languages, key_skills\n\n**Evaluation**: short_summary (max 200 chars), detailed_summary (300-500 words), strengths, weaknesses\n\n**Decision**:\n- **score** (0-100)\n- **recommendation** (REJECT / REVIEW / INTERVIEW / STRONG_HIRE)\n\n### Scoring Grid\n| Score | Decision |\n|-------|-----------|\n| 0-30 | REJECT |\n| 31-60 | REVIEW |\n| 61-80 | INTERVIEW |\n| 81-100 | STRONG_HIRE |" }, "typeVersion": 1 }, { "id": "21753e52-a600-4cc8-9aa8-e442d1ea1d86", "name": "\u2705 Note Validate", "type": "n8n-nodes-base.stickyNote", "position": [ 3104, 528 ], "parameters": { "color": 4, "width": 320, "height": 664, "content": "## \u2705 Validation & Consistency Check\n\nEven though the parser guarantees valid JSON, this node adds a **safety net**:\n\n1. Double-parse if output arrives as a string\n2. Required fields verification\n3. Score clamped to [0, 100]\n4. **Forces consistency** between score \u2194 recommendation (auto-corrects if mismatched, saves original for audit)\n5. Enrichment with Drive metadata, timestamp, fileName\n6. Builds formatted HTML email for recruiter\n\n\ud83d\udca1 Fault-tolerant: errors return `error: true` instead of breaking the flow." }, "typeVersion": 1 }, { "id": "fcab60e1-b45a-4020-8b30-3fe80e62e393", "name": "\ud83d\udcca Note Sheets", "type": "n8n-nodes-base.stickyNote", "position": [ 3392, 1280 ], "parameters": { "color": 2, "width": 320, "height": 440, "content": "## \ud83d\udcca Google Sheets Dashboard\n\nEvery candidate \u2192 one row in the HR sheet.\n\n### Columns\nDate \u00b7 First Name \u00b7 Last Name \u00b7 Email \u00b7 Phone \u00b7 Position \u00b7 Years Exp. \u00b7 Skills \u00b7 **Score** \u00b7 **Recommendation** \u00b7 Summary \u00b7 CV Link\n\n### Tips\n- Conditional formatting on Score column (red/yellow/green)\n- Filter by position\n- Pivot table on Recommendation\n- **Replace `YOUR_SHEET_ID`** with your real Sheet ID\n\n### Alternatives\nNotion DB, Airtable, Baserow, direct PostgreSQL." }, "typeVersion": 1 }, { "id": "1c69aa3a-2daa-4dcc-bb32-a7253312a190", "name": "\ud83c\udfaf Note Notify", "type": "n8n-nodes-base.stickyNote", "position": [ 3952, 448 ], "parameters": { "color": 2, "width": 320, "height": 708, "content": "## \ud83c\udfaf Top Candidate Notification\n\nConfigurable threshold (default: **score \u2265 75**).\n\n### Formatted HTML Email\n- Color-coded header by score (green/blue/yellow/red)\n- KPI cards (score, recommendation, years of experience)\n- Short summary highlighted\n- Candidate profile in table format\n- Skills as badges\n- Strengths/weaknesses in two columns\n- Detailed analysis section\n- Footer with CV link and processing time\n\n### Suggested Extensions\n- Slack/Telegram with action buttons\n- Automatic task creation in Asana/Trello\n- Branch score < 30 \u2192 polite rejection email (with 24-48h delay)" }, "typeVersion": 1 }, { "id": "ebb452bb-16a6-457a-80d4-7e8624914c23", "name": "\u26a0\ufe0f Note Errori", "type": "n8n-nodes-base.stickyNote", "position": [ 3488, 432 ], "parameters": { "color": 6, "width": 320, "height": 616, "content": "## \u26a0\ufe0f Error Handling\n\nEvery failed execution \u2192 automatic email to the Ops team.\n\n### Most Common Causes\n1. CV too long \u2192 Mistral OCR timeout\n2. CV in unsupported language\n3. Corrupted or password-protected PDF\n4. Groq rate limit (free tier: 30 RPM)\n5. JSON schema not respected (rare)\n\n### Production Best Practices\n- Set a global **Error Workflow** in Settings\n- **Retry with backoff** on critical HTTP nodes\n- Log errors to Postgres for analysis\n- Redis dead letter queue for unprocessable CVs" }, "typeVersion": 1 }, { "id": "dcf6f24d-0187-4ae4-a050-10710173ef88", "name": "\ud83d\udd12 Note GDPR", "type": "n8n-nodes-base.stickyNote", "position": [ 800, 1232 ], "parameters": { "color": 6, "width": 340, "height": 468, "content": "## \ud83d\udd12 GDPR & Security\n\nCVs contain **sensitive personal data**.\n\n### Minimum Compliance\n- \u2705 Legal basis documented (Art. 6 GDPR)\n- \u2705 Privacy policy linked in the job posting\n- \u2705 Retention policy: delete rejected CVs after 6-12 months\n- \u2705 Encryption at rest (Docker volumes on encrypted filesystem)\n- \u2705 Access audit log\n- \u2705 Right to erasure (separate workflow)\n\n### Recommended\nA separate scheduled workflow (monthly):\n1. Identify CVs older than X months\n2. Delete files from Drive\n3. Anonymize Sheets row (keep aggregate stats)\n4. Notify DPO" }, "typeVersion": 1 }, { "id": "2430890c-ac15-4f7c-b8cc-f5bf75848c1a", "name": "IMAP Email Trigger (Indeed)", "type": "n8n-nodes-base.emailReadImap", "position": [ 816, 1008 ], "parameters": { "options": { "customEmailConfig": "[\"UNSEEN\"]" }, "postProcessAction": "nothing", "downloadAttachments": true }, "notesInFlow": false, "retryOnFail": true, "typeVersion": 2, "alwaysOutputData": true }, { "id": "4a59236b-8d34-490c-8253-356db1adcd4f", "name": "Extract Email Metadata", "type": "n8n-nodes-base.code", "position": [ 1328, 1008 ], "parameters": { "jsCode": "// \u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\n// EXTRACT EMAIL METADATA FROM INDEED\n// \u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\nconst items = $input.all();\nconst output = [];\n\nfor (const item of items) {\n const subject = item.json.subject || '';\n const from = item.json.from || '';\n const date = item.json.date || new Date().toISOString();\n \n // Regex on Indeed subject: \"New application for X - Y\"\n // \u26a0\ufe0f Adapt this regex to match your Indeed email language/format\n const positionMatch = subject.match(/per\\s+(.+?)\\s+-/i);\n const candidateMatch = subject.match(/-\\s*(.+)$/);\n \n const position = positionMatch ? positionMatch[1].trim() : 'Unknown position';\n const candidateName = candidateMatch ? candidateMatch[1].trim() : 'Unknown candidate';\n \n // Filter valid attachments (PDF/DOCX only)\n const attachments = item.binary || {};\n const attachmentKeys = Object.keys(attachments).filter(k => {\n const mime = attachments[k].mimeType || '';\n return mime.includes('pdf') || mime.includes('word') || mime.includes('document');\n });\n \n if (attachmentKeys.length === 0) {\n continue; // Skip emails without a CV attachment\n }\n \n for (const key of attachmentKeys) {\n const executionStart = Date.now();\n output.push({\n json: {\n _meta: { executionStart },\n subject,\n from,\n date,\n position,\n candidateName,\n attachmentKey: key,\n fileName: attachments[key].fileName || `cv_${Date.now()}.pdf`,\n mimeType: attachments[key].mimeType\n },\n binary: {\n cv: attachments[key]\n }\n });\n }\n}\n\nreturn output;" }, "typeVersion": 2 }, { "id": "34acf446-4393-46e4-9784-f0ed59e3c725", "name": "\u2728 CV Analyzer Chain", "type": "@n8n/n8n-nodes-langchain.chainLlm", "position": [ 2368, 1008 ], "parameters": { "text": "=JOB POSITION APPLIED FOR: {{ $json.position }}\nAPPLICATION DATE: {{ $json.applicationDate }}\nCANDIDATE NAME: {{ $json.candidateName }}\n\nCV TEXT:\n{{ $json.cvText }}\n\nAnalyze the CV and return a structured JSON according to the provided schema.", "batching": {}, "messages": { "messageValues": [ { "message": "You are a senior recruiter specializing in CV evaluation. Your task is to analyze the provided CV and produce a structured, objective, and detailed assessment.\n\nScoring rules:\n- 0-30: profile not aligned, missing key experience/skills\n- 31-60: partially aligned profile, needs review\n- 61-80: good profile, candidate for interview\n- 81-100: excellent profile, strong hire\n\nThe recommendation must be consistent with the score:\n- REJECT (0-30), REVIEW (31-60), INTERVIEW (61-80), STRONG_HIRE (81-100)\n\nBe rigorous but fair. Always explain the reasoning behind the score." } ] }, "promptType": "define", "hasOutputParser": true }, "typeVersion": 1.7 }, { "id": "dca5260c-583a-4f24-ac9c-8e0115ce5a7e", "name": "Structured Output Parser", "type": "@n8n/n8n-nodes-langchain.outputParserStructured", "position": [ 2576, 1248 ], "parameters": { "schemaType": "manual", "inputSchema": "{\n \"type\": \"object\",\n \"properties\": {\n \"nome\": { \"type\": \"string\", \"description\": \"Nome del candidato\" },\n \"cognome\": { \"type\": \"string\", \"description\": \"Cognome del candidato\" },\n \"email\": { \"type\": \"string\", \"description\": \"Email del candidato\" },\n \"telefono\": { \"type\": \"string\", \"description\": \"Numero di telefono\" },\n \"data_candidatura\": { \"type\": \"string\", \"description\": \"Data candidatura YYYY-MM-DD\" },\n \"posizione\": { \"type\": \"string\", \"description\": \"Posizione per cui si candida\" },\n \"anni_esperienza\": { \"type\": \"number\", \"description\": \"Anni di esperienza totale\" },\n \"competenze_chiave\": {\n \"type\": \"array\",\n \"items\": { \"type\": \"string\" },\n \"description\": \"Lista delle competenze tecniche principali\"\n },\n \"lingue\": {\n \"type\": \"array\",\n \"items\": { \"type\": \"string\" },\n \"description\": \"Lingue parlate con livello\"\n },\n \"titolo_studio\": { \"type\": \"string\", \"description\": \"Titolo di studio piu alto\" },\n \"spiegazione_breve\": {\n \"type\": \"string\",\n \"description\": \"Sintesi del profilo in massimo 200 caratteri\"\n },\n \"spiegazione_dettagliata\": {\n \"type\": \"string\",\n \"description\": \"Analisi dettagliata 300-500 parole con punti di forza, gap, fit con la posizione\"\n },\n \"punti_forza\": {\n \"type\": \"array\",\n \"items\": { \"type\": \"string\" },\n \"description\": \"Lista dei principali punti di forza\"\n },\n \"punti_debolezza\": {\n \"type\": \"array\",\n \"items\": { \"type\": \"string\" },\n \"description\": \"Lista delle aree di miglioramento o gap rispetto alla posizione\"\n },\n \"punteggio\": {\n \"type\": \"number\",\n \"minimum\": 0,\n \"maximum\": 100,\n \"description\": \"Punteggio di fit con la posizione da 0 a 100\"\n },\n \"raccomandazione\": {\n \"type\": \"string\",\n \"enum\": [\"REJECT\", \"REVIEW\", \"INTERVIEW\", \"STRONG_HIRE\"],\n \"description\": \"Raccomandazione finale\"\n }\n },\n \"required\": [\n \"nome\",\n \"cognome\",\n \"posizione\",\n \"spiegazione_breve\",\n \"spiegazione_dettagliata\",\n \"punteggio\",\n \"raccomandazione\"\n ]\n}" }, "typeVersion": 1.2 }, { "id": "bb290e42-ed39-4ff0-8bf4-5c4c97469b1f", "name": "\u2705 Validate & Build Email", "type": "n8n-nodes-base.code", "position": [ 2848, 1008 ], "parameters": { "jsCode": "// \u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\n// VALIDATE & ENRICH + BUILD EMAIL HTML\n// \u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\nconst items = $input.all();\nconst output = [];\n\nfor (const item of items) {\n let parsed = item.json.output || item.json;\n \n // Safety net: handle string responses\n if (typeof parsed === 'string') {\n try {\n parsed = JSON.parse(parsed.replace(/```json\\s*/g, '').replace(/```\\s*/g, '').trim());\n } catch (err) {\n output.push({\n json: {\n error: true,\n errorMessage: 'Parse error: ' + err.message,\n rawResponse: item.json\n }\n });\n continue;\n }\n }\n \n // Validate required fields\n const requiredFields = ['nome', 'cognome', 'posizione', 'punteggio', 'raccomandazione'];\n const missing = requiredFields.filter(f => !(f in parsed));\n if (missing.length > 0) {\n output.push({\n json: {\n error: true,\n errorMessage: 'Missing fields: ' + missing.join(', '),\n partialData: parsed\n }\n });\n continue;\n }\n \n // Retrieve upstream context\n const ctx = $node['\ud83e\uddf9 Clean Text'].json;\n const meta = ctx._meta || {};\n \n // Clamp score to [0, 100]\n parsed.punteggio = Math.max(0, Math.min(100, Number(parsed.punteggio) || 0));\n const score = parsed.punteggio;\n \n // Force score <-> recommendation consistency\n const expectedRec = score <= 30 ? 'REJECT'\n : score <= 60 ? 'REVIEW'\n : score <= 80 ? 'INTERVIEW'\n : 'STRONG_HIRE';\n const recCorrected = parsed.raccomandazione !== expectedRec;\n if (recCorrected) {\n parsed.raccomandazione_originale = parsed.raccomandazione;\n parsed.raccomandazione = expectedRec;\n }\n \n // Execution time\n const executionTime = meta.executionStart ? ((Date.now() - meta.executionStart) / 1000).toFixed(1) : 'N/A';\n \n // Score badge color\n const scoreColor = score >= 81 ? '#22c55e'\n : score >= 61 ? '#3b82f6'\n : score >= 31 ? '#f59e0b'\n : '#ef4444';\n \n // Build HTML email\n const puntiForza = (parsed.punti_forza || []).map(p => `
  • ${p}
    • `).join('');\n const puntiDebolezza = (parsed.punti_debolezza || []).map(p => `
  • ${p}
    • `).join('');\n const competenze = (parsed.competenze_chiave || []).map(c => `${c}`).join(' ');\n const lingue = (parsed.lingue || []).join(' \u00b7 ');\n \n const emailHtml = `\n
    \n
    \n
    \ud83c\udfaf New top candidate
    \n
    ${parsed.nome} ${parsed.cognome}
    \n
    ${parsed.posizione}
    \n
    \n \n
    \n \n
    \n
    \n
    Score
    \n
    ${score}/100
    \n
    \n
    \n
    Recommendation
    \n
    ${parsed.raccomandazione}
    \n
    \n
    \n
    Years of experience
    \n
    ${parsed.anni_esperienza || 'N/A'}
    \n
    \n
    \n \n
    \n Summary: ${parsed.spiegazione_breve}\n
    \n \n

    \ud83d\udccb Profile

    \n \n \n \n \n \n
    Email${parsed.email || 'N/A'}
    Phone${parsed.telefono || 'N/A'}
    Education${parsed.titolo_studio || 'N/A'}
    Languages${lingue || 'N/A'}
    \n \n

    \ud83d\udee0 Key Skills

    \n
    ${competenze}
    \n \n
    \n
    \n

    \u2713 Strengths

    \n
      ${puntiForza}
    \n
    \n
    \n

    \u26a0 Weaknesses

    \n
      ${puntiDebolezza}
    \n
    \n
    \n \n

    \ud83d\udcdd Detailed Analysis

    \n
    ${parsed.spiegazione_dettagliata}
    \n \n ${recCorrected ? `
    \u26a0\ufe0f Recommendation auto-corrected from \"${parsed.raccomandazione_originale}\" a \"${parsed.raccomandazione}\" per coerenza con il punteggio.
    ` : ''}\n \n
    \n
    \u23f1 ${executionTime}s
    \n
    \ud83d\udcc4 Open CV on Drive
    \n
    \ud83e\udd16 Llama 3.3 70B (Groq)
    \n
    \n
    \n
    \n `;\n \n output.push({\n json: {\n ...parsed,\n _emailHtml: emailHtml,\n _emailSubject: `\ud83c\udfaf ${parsed.nome} ${parsed.cognome} - Score ${score}/100 (${parsed.raccomandazione})`,\n cv_drive_link: ctx.driveLink,\n cv_filename: ctx.fileName,\n processed_at: new Date().toISOString(),\n execution_time: executionTime,\n error: false\n }\n });\n}\n\nreturn output;" }, "typeVersion": 2 }, { "id": "093a5c93-4f3a-4572-b6a6-e4a1918584fc", "name": "Has Error?", "type": "n8n-nodes-base.if", "position": [ 3152, 1008 ], "parameters": { "options": {}, "conditions": { "options": { "version": 1, "leftValue": "", "caseSensitive": true, "typeValidation": "strict" }, "combinator": "and", "conditions": [ { "id": "error-check", "operator": { "type": "boolean", "operation": "equals" }, "leftValue": "={{ $json.error }}", "rightValue": true } ] } }, "typeVersion": 2 }, { "id": "8c1f4352-064f-420e-8b24-68c4a127fea1", "name": "Score >= 75?", "type": "n8n-nodes-base.if", "position": [ 3728, 1104 ], "parameters": { "options": {}, "conditions": { "options": { "version": 1, "leftValue": "", "caseSensitive": true, "typeValidation": "strict" }, "combinator": "and", "conditions": [ { "id": "score-threshold", "operator": { "type": "number", "operation": "gte" }, "leftValue": "={{ $json.punteggio }}", "rightValue": 75 } ] } }, "typeVersion": 2 }, { "id": "1b31fddc-ba98-462c-b85e-e67440441926", "name": "\ud83d\udce7 Notify Recruiter", "type": "n8n-nodes-base.emailSend", "position": [ 4048, 992 ], "parameters": { "html": "={{ $json._emailHtml }}", "options": {}, "subject": "={{ $json._emailSubject }}", "toEmail": "user@example.com", "fromEmail": "user@example.com" }, "typeVersion": 2.1 }, { "id": "364cbfe9-aedc-4bae-b67f-26f371aa86a0", "name": "\u26a0\ufe0f Notify Error", "type": "n8n-nodes-base.emailSend", "position": [ 3600, 864 ], "parameters": { "options": {}, "subject": "\u274c CV Rating Pipeline Error", "toEmail": "user@example.com", "fromEmail": "user@example.com" }, "typeVersion": 2.1 }, { "id": "8a1e7307-e39c-4830-9b27-45a857d79598", "name": "Share file", "type": "n8n-nodes-base.googleDrive", "position": [ 1696, 1008 ], "parameters": { "fileId": { "__rl": true, "mode": "id", "value": "={{ $json.id }}" }, "options": {}, "operation": "share", "permissionsUi": { "permissionsValues": { "role": "reader", "type": "anyone" } } }, "typeVersion": 3 }, { "id": "e74acfd0-5f82-4ffa-840d-113b9d6e3c52", "name": "\ud83d\udcc1 Upload Drive", "type": "n8n-nodes-base.googleDrive", "position": [ 1536, 1008 ], "parameters": { "name": "={{ $json.fileName }}", "driveId": { "__rl": true, "mode": "list", "value": "", "cachedResultUrl": "", "cachedResultName": "" }, "options": {}, "folderId": { "__rl": true, "mode": "id", "value": "", "cachedResultUrl": "", "cachedResultName": "" }, "inputDataFieldName": "cv" }, "typeVersion": 3 }, { "id": "0b9ce595-7d22-46c9-80b9-cec8f930a152", "name": "\ud83e\udde0 Mistral OCR", "type": "n8n-nodes-base.httpRequest", "position": [ 1936, 1008 ], "parameters": { "url": "https://api.mistral.ai/v1/ocr", "method": "POST", "options": { "timeout": 60000 }, "jsonBody": "={\n \"model\": \"mistral-ocr-latest\",\n \"document\": {\n \"type\": \"document_url\",\n \"document_url\": \"https://drive.google.com/uc?export=download&id={{ $node['\ud83d\udcc1 Upload Drive'].json.id }}\"\n },\n \"include_image_base64\": false\n}", "sendBody": true, "sendHeaders": true, "specifyBody": "json", "authentication": "predefinedCredentialType", "headerParameters": { "parameters": [ { "name": "Content-Type", "value": "application/json" } ] }, "nodeCredentialType": "mistralCloudApi" }, "typeVersion": 4.2 }, { "id": "1e3d73b1-f241-49da-be52-8b44d3e54f1a", "name": "Groq Chat Model", "type": "@n8n/n8n-nodes-langchain.lmChatGroq", "position": [ 2320, 1248 ], "parameters": { "model": "llama-3.3-70b-versatile", "options": { "temperature": 0.2, "maxTokensToSample": 2000 } }, "typeVersion": 1 }, { "id": "98f7dfd0-d690-4fdc-9ddd-ef3e15fc3bdb", "name": "Append row in sheet", "type": "n8n-nodes-base.googleSheets", "position": [ 3488, 1120 ], "parameters": { "operation": "append", "sheetName": { "__rl": true, "mode": "list", "value": "", "cachedResultUrl": "", "cachedResultName": "" }, "documentId": { "__rl": true, "mode": "list", "value": "", "cachedResultUrl": "", "cachedResultName": "" } }, "typeVersion": 4.7 }, { "id": "36269d3c-8615-4cc9-b84e-146f89a92ddb", "name": "CV check", "type": "n8n-nodes-base.if", "position": [ 1168, 1104 ], "parameters": { "options": {}, "conditions": { "options": { "version": 1, "leftValue": "", "caseSensitive": true, "typeValidation": "strict" }, "combinator": "and", "conditions": [ { "id": "error-check", "operator": { "type": "string", "operation": "contains" }, "leftValue": "={{ $json.subject }}", "rightValue": "ADD_HEADER_INDEED_MAIL" } ] } }, "typeVersion": 2 }, { "id": "a5720f62-a4e6-4737-a4c8-e2ae8b914b4a", "name": "\ud83e\uddf9 Clean Text", "type": "n8n-nodes-base.code", "position": [ 2128, 1008 ], "parameters": { "jsCode": "// \u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\n// CLEAN & NORMALIZE OCR OUTPUT\n// \u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\u2550\nconst items = $input.all();\nconst output = [];\n\nfor (const item of items) {\n const ocrResponse = item.json;\n let fullText = '';\n \n if (ocrResponse.pages && Array.isArray(ocrResponse.pages)) {\n fullText = ocrResponse.pages\n .map(p => p.markdown || p.text || '')\n .join('\\n\\n');\n }\n \n // Normalize whitespace and newlines\n fullText = fullText\n .replace(/\\s{3,}/g, ' ')\n .replace(/\\n{3,}/g, '\\n\\n')\n .trim();\n \n // Retrieve context from previous nodes (production workflow)\n const meta = $('Extract Email Metadata').first().json;\n const drive = $('\ud83d\udcc1 Upload Drive').first().json;\n \n output.push({\n json: {\n _meta: meta._meta,\n cvText: fullText,\n candidateName: meta.candidateName,\n position: meta.position,\n applicationDate: meta.date,\n fileName: meta.fileName,\n driveLink: drive.webViewLink,\n driveId: drive.id,\n textLength: fullText.length\n }\n });\n}\n\n// Drop binary data to keep the flow lightweight\nreturn output.map(item => ({ json: item.json }));" }, "typeVersion": 2 }, { "id": "842769b0-4d36-4bda-a591-75955f6df68c", "name": "\u2699\ufe0f Note Configuration", "type": "n8n-nodes-base.stickyNote", "position": [ 400, 112 ], "parameters": { "color": 3, "width": 380, "height": 460, "content": "## \u2699\ufe0f Configuration\n\nAll user-configurable variables are grouped here.\n**Edit this node before activating the workflow.**\n\n| Variable | Description |\n|---|---|\n| `DRIVE_ROOT_FOLDER_ID` | Google Drive folder ID where CVs are saved |\n| `SHEET_ID` | Google Sheets ID for the HR dashboard |\n| `RECRUITER_EMAIL` | Email address to notify for top candidates |\n| `OPS_EMAIL` | Email address to notify on errors |\n| `SENDER_EMAIL` | Sender address (must match your SMTP account) |\n| `SCORE_THRESHOLD` | Minimum score to trigger recruiter notification (default: 75) |\n| `IMAP_SUBJECT_FILTER` | Subject keyword to filter Indeed emails |\n\n\ud83d\udca1 Folder/Sheet IDs are found in the URL of the respective Google resource." }, "typeVersion": 1 } ], "active": false, "settings": { "binaryMode": "separate", "executionOrder": "v1" }, "versionId": "d47e3399-24aa-43c0-a5cc-7208a3e64444", "connections": { "CV check": { "main": [ [ { "node": "Extract Email Metadata", "type": "main", "index": 0 } ] ] }, "Has Error?": { "main": [ [ { "node": "\u26a0\ufe0f Notify Error", "type": "main", "index": 0 } ], [ { "node": "Append row in sheet", "type": "main", "index": 0 } ] ] }, "Share file": { "main": [ [ { "node": "\ud83e\udde0 Mistral OCR", "type": "main", "index": 0 } ] ] }, "Score >= 75?": { "main": [ [ { "node": "\ud83d\udce7 Notify Recruiter", "type": "main", "index": 0 } ] ] }, "Groq Chat Model": { "ai_languageModel": [ [ { "node": "\u2728 CV Analyzer Chain", "type": "ai_languageModel", "index": 0 } ] ] }, "\ud83e\uddf9 Clean Text": { "main": [ [ { "node": "\u2728 CV Analyzer Chain", "type": "main", "index": 0 } ] ] }, "\ud83e\udde0 Mistral OCR": { "main": [ [ { "node": "\ud83e\uddf9 Clean Text", "type": "main", "index": 0 } ] ] }, "\ud83d\udcc1 Upload Drive": { "main": [ [ { "node": "Share file", "type": "main", "index": 0 } ] ] }, "Append row in sheet": { "main": [ [ { "node": "Score >= 75?", "type": "main", "index": 0 } ] ] }, "\u2699\ufe0f Configuration": { "main": [ [ { "node": "CV check", "type": "main", "index": 0 } ] ] }, "\u2728 CV Analyzer Chain": { "main": [ [ { "node": "\u2705 Validate & Build Email", "type": "main", "index": 0 } ] ] }, "Extract Email Metadata": { "main": [ [ { "node": "\ud83d\udcc1 Upload Drive", "type": "main", "index": 0 } ] ] }, "Structured Output Parser": { "ai_outputParser": [ [ { "node": "\u2728 CV Analyzer Chain", "type": "ai_outputParser", "index": 0 } ] ] }, "\u2705 Validate & Build Email": { "main": [ [ { "node": "Has Error?", "type": "main", "index": 0 } ] ] }, "IMAP Email Trigger (Indeed)": { "main": [ [ { "node": "\u2699\ufe0f Configuration", "type": "main", "index": 0 } ] ] } } }