{
  "name": "UK University Data Pipeline - Template",
  "nodes": [
    {
      "id": "1",
      "name": "Start",
      "type": "n8n-nodes-base.manualTrigger",
      "position": [
        200,
        250
      ],
      "typeVersion": 1
    },
    {
      "id": "2",
      "name": "Fetch UCAS / Program Page",
      "type": "n8n-nodes-base.httpRequest",
      "position": [
        420,
        250
      ],
      "parameters": {
        "url": "https://example.com/api-or-program-url",
        "options": {
          "method": "GET"
        }
      },
      "typeVersion": 1
    },
    {
      "id": "3",
      "name": "Preprocess HTML / JSON",
      "type": "n8n-nodes-base.function",
      "position": [
        660,
        250
      ],
      "parameters": {
        "functionCode": "const raw = items[0].json;\n// Clean HTML/JSON before LLM\nlet html = raw.body || raw.html || raw;\nif (typeof html === 'string') {\n  html = html\n    .replace(/<script[\\s\\S]*?<\\/script>/gi, '')\n    .replace(/<style[\\s\\S]*?<\\/style>/gi, '')\n    .replace(/\\s+/g, ' ');\n}\nreturn [{ json: { cleaned: html } }];"
      },
      "typeVersion": 1
    },
    {
      "id": "4",
      "name": "OpenAI Extract (LLM)",
      "type": "n8n-nodes-base.openAi",
      "position": [
        900,
        250
      ],
      "parameters": {
        "model": "gpt-4o-mini",
        "messages": [
          {
            "role": "system",
            "content": "Extract structured program data from HTML. Return ONLY valid JSON."
          },
          {
            "role": "user",
            "content": "={{ 'HTML INPUT: ' + $json.cleaned }}"
          }
        ],
        "operation": "chatCompletion",
        "temperature": 0
      },
      "typeVersion": 2
    },
    {
      "id": "5",
      "name": "Parse LLM JSON",
      "type": "n8n-nodes-base.function",
      "position": [
        1150,
        250
      ],
      "parameters": {
        "functionCode": "let output = items[0].json;\nlet text = output.choices?.[0]?.message?.content || null;\nlet parsed;\ntry {\n  parsed = JSON.parse(text);\n} catch (e) {\n  parsed = { error: 'INVALID_JSON_FROM_LLM', raw: text };\n}\nreturn [{ json: parsed }];"
      },
      "typeVersion": 1
    },
    {
      "id": "6",
      "name": "Save to Postgres",
      "type": "n8n-nodes-base.postgres",
      "position": [
        1400,
        250
      ],
      "parameters": {
        "query": "INSERT INTO programs(data) VALUES(:data::jsonb);",
        "options": {
          "queryParams": "= { \"data\": JSON.stringify($json) }"
        },
        "operation": "executeQuery",
        "additionalFields": {}
      },
      "typeVersion": 1
    }
  ],
  "active": false,
  "versionId": "1.0.0",
  "connections": {
    "Start": {
      "main": [
        [
          {
            "node": "Fetch UCAS / Program Page",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Parse LLM JSON": {
      "main": [
        [
          {
            "node": "Save to Postgres",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "OpenAI Extract (LLM)": {
      "main": [
        [
          {
            "node": "Parse LLM JSON",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Preprocess HTML / JSON": {
      "main": [
        [
          {
            "node": "OpenAI Extract (LLM)",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Fetch UCAS / Program Page": {
      "main": [
        [
          {
            "node": "Preprocess HTML / JSON",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  }
}