AI News Classifier for TechCrunch Articles - n8n Workflow

• AI Researchers & Analysts: Who need curated, scored, and summarized news about Artificial Intelligence.


• Content Curators: Individuals or teams looking to automate the discovery and analysis of niche content.


• n8n Users: Developers seeking advanced examples of combining web scraping, LLM agents, and structured output parsing within an n8n node setup.


• Data Journalists: Who require a structured data feed of high-quality AI news.

Keeping up with the fast-paced world of Artificial Intelligence requires robust automation. This specific n8n workflow solves the tedious task of manually monitoring TechCrunch’s AI category. It automates the entire pipeline, acting as a personal, intelligent news filtering agent.

The core value of this n8n template lies in its use of an AI Agent powered by GPT-4.1-nano. Not only does it scrape the full text, but it also applies a complex scoring rubric (1-100) and categorizes the content, all delivered in a structured Thai format. Every processed item is logged reliably into a Google Sheet and pushed instantly as a notification via Telegram. This robust n8n node combination ensures high efficiency and data quality.

This automation starts with a daily recurring n8n trigger and follows these nine steps:

1. Scheduled Activation: The Schedule Trigger n8n node fires every day at 6 AM, initiating the news collection process.


2. Initial Scraping: An HTTP Request n8n node fetches the main TechCrunch AI category page HTML content.


3. Metadata Extraction: The HTML n8n node extracts all available article URLs and their publication dates using specific CSS selectors.


4. Filtering & Formatting: The flow uses a Set n8n node to format dates and a Switch n8n node (SelectDate) to filter the results, ensuring only articles published yesterday are processed.


5. Rate Limiting: A Code n8n node introduces a random delay (4 to 10 seconds) before proceeding, enforced by a Wait n8n node, mitigating the risk of being blocked during subsequent individual article scraping.


6. Full Content Fetch: An HTTP Request n8n node fetches the full content of the filtered article URL, and a subsequent HTML n8n node extracts the main article body.


7. AI Analysis: The crucial step utilizes an advanced AI Agent n8n node paired with the GPT-4.1-nano LLM and a Structured Output Parser. The agent uses a detailed Thai prompt to summarize the news, critically score it based on impact and quality, and assign it a single category, enforcing a strict JSON output schema.


8. Data Logging: The enriched, structured data (headline, score, summary, category, URL, date) is then appended to a specified Google Sheet using the Google Sheets n8n node.


9. Instant Notification: Finally, a rich, Thai-language summary of the analyzed news is sent instantly to a configured chat via the Telegram n8n node.

To deploy this comprehensive n8n workflow, follow these steps:

1. Import the n8n Workflow: Copy the provided JSON code and paste it directly into your n8n instance using the 'New' -> 'Import from JSON' function.


2. Configure Credentials:

1. Adjust Filtering: Review the SelectDate n8n node. Currently, it filters for articles exactly one day old. Adjust the date expression if you wish to scrape articles from a different time range.


2. Activate: Save and activate the n8n workflow. It will automatically execute daily.

Schedule Trigger (n8n trigger):
Function: Initiates the entire n8n workflow on a fixed schedule.
Key Configuration: Set to trigger daily at 6 AM (triggerAtHour: 6).
TechCrunchNews-AI1 (HTTP Request n8n node):
Function: Fetches the raw HTML of the target category page.
Key Configuration: URL is set to https://techcrunch.com/category/artificial-intelligence/.
GetDate1 (HTML n8n node):
Function: Performs CSS selection to extract structured data (URLs and Publication Dates) from the scraped HTML.
Key Configuration: Extracts datetime attribute from .loop-cardmeta > time and data-destinationlink from .loop-cardtitle > a.
SelectDate (Switch n8n node):
Function: Filters items, ensuring that only articles published yesterday are passed for processing.
Key Configuration: Compares DatePublished with {{ $now.minus(1,'days').format('dd/MM/yyyy') }}.
TimeWaitRandom1 (Code n8n node):
Function: Generates a random time delay (4 to 10 seconds) to ensure responsible web scraping practices.
GetBody (HTML n8n node):
Function: Extracts the full article text from the specific article URL.
Key Configuration: Uses the CSS selector div[class="entry-content"] to isolate the main article body.
AI Agent1 (AI Agent n8n node):
Function: The central intelligence unit. It processes the article body using a sophisticated Thai prompt to generate a summary, calculate a quality score (1-100), and categorize the content.
Key Configuration: Uses the Gpt-4.1-nano language model and enforces output structure using the connected parser.
NewsData (Google Sheets n8n node):
Function: Persists the final analyzed data into a Google Sheet for long-term storage and analysis.
Key Configuration: Operation: append, mapping six key fields including score, headline, and summary.
Send a text message (Telegram n8n node):
Function: Delivers immediate alerts containing the analyzed news summary to the target Telegram chat.