Build RSS to Notion Knowledge Base with Make.com

Overview

This is a fully automated knowledge archival system designed for advanced users.

By connecting multiple tools, it transforms fragmented RSS subscriptions into a structured, searchable knowledge base:

1. RSS Monitoring - Capture latest articles
2. AI Processing - OpenAI translates titles and formats JSON
3. Data Cleaning - Regex truncates title length
4. Format Conversion - HTML to Notion-friendly Markdown
5. PDF Generation - Web pages to PDF permanent backup
6. Cloud Archival - Upload to Google Drive
7. Knowledge Base - Create Notion entries

End-to-end automation from RSS to Notion

---

Core Decision Factors

When building a knowledge management automation system, consider:

• Configuration Complexity - Do you need to handle complex API authentication (like OAuth)?
• Integration Depth - Can it handle rich media and format conversion?
• Fault Tolerance - Will it crash when source data is missing or AI output is incorrect?
• Search Capability - Is archived content easy to search through?
• Operating Cost - Are API call costs proportional to time saved?

---

Technical Specifications

Specification	Value	Notes
AI Engine	GPT-3.5 Turbo	For title translation and data cleaning
Token Limit	1000	Maximum consumption per run
PDF Engine	PDF.co	Web page to PDF conversion
Link Validity	60 minutes	PDF.co temporary links
Notion Title Limit	~100 characters	Recommend truncating to 90 characters
Storage Solution	Google Drive + Notion	Drive for PDFs, Notion for metadata

---

Prerequisites

Before starting, ensure you have:

• Make.com account (free registration)
• OpenAI API key
• PDF.co account
• Google Cloud Platform account (for Drive API)
• Notion account and database

Notion Database Structure

Database field setup

Create an RSS archival database with the following fields:

• Title (Title) - Article title (English)
• Original Link (URL) - RSS article link
• PDF Link (URL) - Google Drive backup link
• Published (Date) - Article publication date
• Created (Date) - Archival timestamp

---

Workflow Architecture

Stage 1: RSS Monitoring and AI Processing

Prompt design for AI to output structured JSON

RSS Monitoring Configuration:

• Add RSS Watch module
• Configure subscription feed URL
• Set check frequency

OpenAI Processing:

• Translate titles to English (avoid filename encoding issues)
• Output formatted JSON structure

Prompt Example:

Please translate the following title to English and return in JSON format:
{"title": "translated English title"}

Original title: {{rss_title}}

---

Stage 2: Data Cleaning

Using regex to truncate title to first 90 characters

Handle potential issues with AI output:

JSON Parsing:

• OpenAI output may not be standard JSON
• Sometimes adds extra commas or periods
• Requires JSON Parser module

Title Truncation:

• Notion title limit is approximately 100 characters
• Use regex to truncate to first 90 characters
• Avoid errors from overly long titles

---

Stage 3: Format Conversion

Convert HTML content to Markdown:

Why Convert?

• Markdown is Notion’s native format
• Conversion enables full-text search
• Preserves article structure and formatting

Considerations:

• Different RSS feeds have different structures
• Some have content in description, others in content
• Need to check each source individually

---

Stage 4: PDF Generation and Cloud Archival

Use PDF.co to convert web pages to PDF:

Configuration Points:

• Input original article URL
• Obtain temporary PDF link
• Link valid for only 60 minutes

Google Drive Upload:

GCP project configuration interface

This is the most complex part of the entire workflow:

1. Go to Google Cloud Platform
2. Create a new project
3. Configure OAuth consent screen
4. Enable Drive API
5. Generate Client ID and Secret
6. Complete authorization in Make

Warning: This configuration process is the biggest barrier for non-technical users.

---

Stage 5: Notion Write

Write all data to Notion:

Content Written:

• Translated English title
• Original article link
• Google Drive PDF link
• Publication date
• Markdown body (appended to page)

---

Pitfalls to Avoid

Common issues during implementation:

1. Google Drive Configuration Complexity - Requires creating GCP project, configuring OAuth, enabling API - a tedious process

2. Notion Title Character Limit - Exceeding 100 characters causes errors, must use regex truncation

3. Unstable JSON Parsing - AI output format may have issues, need good error handling

4. Inconsistent RSS Feed Structure - Different feeds have fields in different locations, need to check each one

5. Short PDF Link Validity - Only 60 minutes, must promptly upload to Drive

---

Use Cases

Recommended For

• Information Collectors/Researchers - Need permanent archival of massive industry updates
• Heavy Notion Users - Want to centrally manage all reading materials
• Make.com Learners - Excellent intermediate to advanced practice project

May Not Suit

• Casual news readers (traditional RSS readers like Feedly offer better experience)
• Complete beginners (GCP configuration is extremely tedious)
• Cost-sensitive users (API calls accumulate costs)

---

FAQ

Why convert HTML to Markdown?

Converting to Markdown makes content native to Notion, enabling full-text search within Notion and dramatically improving retrieval efficiency.

Is Google Drive configuration complex?

Yes, it requires creating a project in Google Cloud Platform, configuring OAuth, and enabling Drive API. This is the highest technical barrier in the entire workflow.

Does Notion have title character limits?

Yes, approximately 100 characters. If AI-translated titles exceed this, errors occur. Use regex to truncate to the first 90 characters.

Do PDF links expire?

Yes, PDF.co generates links valid for only 60 minutes. You must promptly upload to Google Drive for permanent storage.

---

Next Steps

After mastering the basics, you can try:

• Add more RSS subscription feeds
• Integrate AI summarization functionality
• Add automatic tag categorization
• Set up email notifications for important articles

Feel free to leave comments with questions!