Daily Tech News Curation and AI Summarization using Mistral - n8n Workflow

Developers, researchers, or analysts who need a daily, filtered digest of technological, scientific, and applied math news.
Users looking for advanced examples of n8n templates integrating multiple RSS sources.
n8n users interested in deploying powerful AI agents for content analysis and summarization.
Professionals wanting a hands-on, customizable n8n workflow utilizing the Mistral Cloud chat model.

This powerful n8n workflow solves the problem of information overload by automating the entire news curation process. Instead of manually checking dozens of RSS feeds, this solution fetches news from nine different technology and science sources (like ScienceDaily, MIT Technology Review, and The Hacker News), standardizes the data, and applies strict filtering rules (e.g., filtering by date and keywords like 'porn' or 'layoff'). The core value is delivered by the AI Agent, which first analyzes article titles to select the ten most relevant stories and then generates concise, short summaries for those selected articles. This customized n8n workflow acts as a personalized technical newspaper, ensuring you only receive highly relevant, AI-processed insights. Using Mistral as the LLM ensures flexibility and performance within this complex n8n automation.

The n8n workflow begins with the Schedule n8n trigger, configured to run every weekday.

1. News Gathering: The schedule n8n trigger initiates multiple RSS Feed Read n8n nodes (ScienceDaily, MIT Technology Review, etc.) and a DuckDuckGo Search n8n node to gather a large list of recent news articles.


2. Data Standardization: Each source's output is processed by a Filter Fields n8n node to ensure a consistent data structure (title, pubDate, link, content). For sources that lack content (like lobste.rs or DuckDuckGo), an HTTP Request n8n node is used, followed by Markdown Conversion and a Rejoin Data n8n node to scrape the link content.


3. Data Merging and Filtering: All standardized data streams are combined using a Merge n8n node. A Filter by Datetime n8n node ensures only fresh, daily news is kept, followed by the Remove Certain Content n8n node which uses regex to eliminate articles containing specified keywords (e.g., opinion, gossip).


4. AI Analysis (Top 10 Selection): The remaining clean data is aggregated. The News Analyzer AI Agent n8n node, powered by the Mistral Cloud Chat Model and constrained by a JSON Output Parser n8n node, analyzes all titles. It summarizes the day's news and intelligently selects the top 10 articles, prioritizing tech, science, and math research, while ensuring no more than one article per source website. This specific n8n node setup is crucial for accuracy.


5. Summarization Preparation: The Split Out News Data n8n node separates the selected top 10 articles. This limited list is merged (Join Outputs n8n node) back with the original, scraped content data from the initial ingestion phase.


6. AI Summarization (Content Generation): The Website Summarizer AI Agent n8n node, also using the Mistral Cloud Chat Model and a Structured JSON Output Parser n8n node, takes the full content of the top 10 articles and generates a concise, descriptive one-sentence summary for each.


7. Final Output: A Final Aggregate n8n node collects the summarized top 10 news items and the daily summary generated earlier, providing the complete curated digest. This robust n8n workflow ensures high-quality, relevant results.

To set up this powerful n8n workflow, follow these steps:

1. Import the n8n template: Copy the provided JSON code and import it into your n8n instance via the Workflows section.


2. Install Custom Node (If needed): The n8n workflow utilizes the n8n-nodes-duckduckgo-search custom n8n node. If you are self-hosting, ensure this node is installed.


3. Configure Credentials:

1. Review Filtering Logic: Customize the Remove Certain Content Filter n8n node to adjust the keywords used to exclude unwanted news based on your preferences. You can also adjust the conditions on the Filter by Datetime n8n node.


2. Review Sub-Workflow: Note that the "Call Get Webpage Content to Data" n8n node requires a separate Sub-Workflow to execute properly. Ensure the linked workflow (ID: MqafS0zommXM7soB) is installed and functional.


3. Activate and Test: Enable the n8n workflow. It uses a Schedule n8n trigger, so run it once manually to ensure all connections and the complex AI agents are configured correctly.

This n8n workflow relies on several specialized and core n8n node types:

Schedule (n8n trigger): The primary n8n trigger that starts the automation.
Function: Runs the n8n workflow daily on weekdays after 12:00 PM to gather the latest news.
RSS Feed Read (n8n node): Multiple instances (e.g., ScienceDaily Feed, MIT Technology Review Feed) gather initial news links and titles.
Function: Fetches structured data from specific news RSS feeds.
HTTP Request (n8n node) & Markdown Conversion (n8n node): Used sequentially to fetch and clean the raw webpage content from news links when the RSS feed doesn't provide the article body.
Merge & Aggregate (n8n nodes): Core flow control n8n nodes used to combine multiple disparate data streams and prepare a single item containing the compiled list of news for the AI Agent.
Filter by Datetime & Remove Certain Content (n8n nodes): Critical data cleaning steps.
Key Configuration: Filters out news older than the current day and excludes articles whose titles, links, or content match regex patterns like opinion|rumour|layoff|porn.
News Analyzer (AI Agent n8n node): The first stage of AI processing in this n8n workflow.
Function: Receives all gathered news titles and uses the LLM (Mistral) to generate a high-level summary and select the top 10 articles based on strict criteria.
Key Configuration: Uses Mistral Cloud and is strictly governed by the JSON Output Parser n8n node to ensure a structured response.
Website Summarizer (AI Agent n8n node): The second stage of AI processing.
Function: Takes the content of the top 10 selected articles and generates a brief, concise, one-sentence summary for each.
Mistral Cloud Chat Model (n8n node): The selected Large Language Model used by both AI Agent n8n nodes.
* Key Configuration: Uses mistral-small-latest model.