Building an AI News Automation System from Scratch: A Hands-On Tutorial for Deploying n8n + DeepSeek + Hugo on Ubuntu

Author: HappyRock

Target Audience: Developers, Indie Developers, Technical Managers, AI Enthusiasts

Tech Stack: Ubuntu + Docker + n8n + DeepSeek + Hugo

Preface

In the past few years, we’ve been used to:

Open RSS
↓
Browse News
↓
Filter Hot Topics
↓
Summarize
↓
Write WeChat Article
↓
Publish Blog Post

The whole process often takes 1–2 hours.

But in the era of AI Agents, these tasks can already be handed over to AI.

This article will document the complete process from scratch:

Deploy n8n on Ubuntu
Configure Docker Compose
Connect to DeepSeek API
Automatically fetch AI news
Automatically filter and rank hot topics
Automatically generate WeChat public account content
Automatically output Hugo Markdown

The result will be your own AI content production system.

1. Why n8n?

When many people first encounter automation, they run into tools like:

Python + Crontab Shell Script Airflow Jenkins

These solutions, while powerful, have a higher barrier to entry.

n8n’s advantages are:

Low‑code
Visual workflow designer
AI support
Webhook support
API support
Database support
Workflow orchestration

Essentially:

Trigger
   ↓
Input
   ↓
Process
   ↓
Output

It’s no different in principle from what programmers are familiar with:

#php

$data = input();

$result = process($data);

output($result);

2. Deploy Docker on Ubuntu

First, confirm your system:

#bash

lsb_release -a

Example output:

Ubuntu 20.04 LTS

Install Docker:

#bash

curl -fsSL https://get.docker.com | bash

Check the version:

#bash

docker --version

Example:

Docker version 28.1.1

3. Verify Docker Compose

Newer Docker versions include Compose built‑in.

Check:

#bash

docker compose version

Example:

Docker Compose version v5.0.2

If you see:

docker-compose: command not found

it means you are using an older version.

4. Understanding Docker Compose’s Working Directory

Many beginners run into:

#bash

docker compose up -d

and get an error:

no configuration file provided: not found

Reason:

Docker Compose requires a file named

docker-compose.yml

in the current directory.

For example:

#bash

pwd

returns

/home/ubuntu/n8n

and that directory must contain

docker-compose.yml

5. Deploy n8n

Create the directory:

#bash

mkdir ~/n8n

cd ~/n8n

Create the compose file:

#bash

nano docker-compose.yml

Content:

#yaml

version: '3'

services:
  n8n:
    image: n8nio/n8n:latest
    container_name: n8n

    ports:
      - "5678:5678"

    environment:
      - TZ=Asia/Shanghai

    volumes:
      - ./n8n_data:/home/node/.n8n

    restart: always

Start the container:

#bash

docker compose up -d

Verify:

#bash

docker ps

If you see

n8n

the deployment succeeded.

6. First Access to n8n

Open your browser and go to:

http://<your_server_ip>:5678

On first access you will see:

Set up owner account

Fill in:

Email
Password
First Name
Last Name

to create the admin account.

7. Understanding Core n8N Nodes

I recommend mastering these first:

Trigger

Triggers:

Manual Trigger
Schedule Trigger
Webhook

Set

Set variables, e.g.:

{
  "name":"Jack"
}

IF

Conditional branching:

if
else

HTTP Request

Call APIs:

DeepSeek
OpenAI
Enterprise systems

Code

JavaScript node:

#JavaScript

return [{
  json:{
    message:"hello"
  }
}]

Merge

Merge data streams.

8. Build Your First AI News System

Goal:

RSS
↓
DeepSeek
↓
AI Daily Report

Step1: Schedule Trigger

Configuration:

Every Day
08:00

Step2: RSS Feed Read

OpenAI:

https://openai.com/news/rss.xml

Anthropic:

https://www.anthropic.com/news/rss.xml

DeepMind:

https://deepmind.google/blog/rss.xml

Step3: Merge

Mode:

Append

Effect:

OpenAI + Anthropic + DeepMind = Unified news stream

Step4: Code Node to Format News

#JavaScript

const news = items.map(i => ({
  title: i.json.title,
  link: i.json.link
}));

return [{
  json:{
    news
  }
}];

9. Call DeepSeek

HTTP Request node:

POST
https://api.deepseek.com/chat/completions

Headers:

Authorization
Bearer sk-xxxx

Content-Type
application/json

10. Common Error Troubleshooting

1. Authorization failed

Reasons:

Wrong Bearer format
Invalid API key
Header placed incorrectly

Correct format:

Authorization: Bearer sk-xxxx

2. Invalid JSON

Example error:

Bad control character

Reason:

Line breaks break JSON

Solution:

#JavaScript

JSON.stringify()

3. [object Object]

Error:

The value "[object Object]" is not supported

Reason: Trying to concatenate an object directly into a string.

Wrong:

#JavaScript

content: $json.news

Correct:

#JavaScript

content: JSON.stringify($json.news)

11. Enterprise‑Grade Prompt Design

Ordinary user prompt:

Please summarize the following news

Average results.

Enterprise‑grade prompt:

You are a senior AI industry analyst and tech media editor.

Tasks:

1. Value scoring
2. News filtering
3. Hot topic ranking
4. Categorization
5. Output a WeChat‑style article

12. AI News Filtering System

Let DeepSeek perform:

Filter
↓
Rank
↓
Analyze
↓
Output

Requirement:

Score >= 7

Output:

</>MarkDown

# Today's AI Industry Updates

## Key Highlights

## Industry Analysis

## Takeaways for Developers

## Summary

13. Upgrade to a Content Factory

Add further automation:

Auto‑generated titles

Prompt:

Generate 10 viral WeChat article titles

Example:

OpenAI drops another bombshell – is the AI industry about to change?

Auto‑generated cover copy

Example:

OpenAI and Anthropic make moves on the same day – AI competition enters a new phase

Auto‑generated Markdown

Output:

</>MarkDown

---
title: Today's AI Industry Updates
date: 2026-06-15
draft: false
---

Article content...

14. Integrate with Hugo for Automatic Publishing

Workflow:

RSS
↓
DeepSeek
↓
Markdown
↓
Write File
↓
Git Push
↓
GitHub Action
↓
Hugo
↓
HappyRock Blog

Automatically creates:

content/blog/2026-06-15-ai-news.md

15. Future Upgrade Roadmap

Level 1

RSS
↓
Summary

Level 2

RSS
↓
Filter
↓
Rank

Level 3

RSS
↓
WeChat article

Level 4

RSS
↓
AI Agent
↓
WeChat public account
↓
Blog
↓
Email
↓
Corporate knowledge base

Summary

Through this hands‑on project, we have successfully:

✅ Deployed Docker on Ubuntu
✅ Deployed n8n with Docker Compose
✅ Configured the DeepSeek API
✅ Aggregated multiple RSS sources
✅ Filtered and ranked AI news
✅ Generated WeChat‑style content
✅ Designed an automatic Hugo publishing architecture

More importantly, we have grasped a key concept:

n8n is not just a simple automation tool – it is a workflow orchestration platform for the AI Agent era.

For developers, the truly valuable skills in the future will no longer be just writing code, but being able to leverage AI, workflows, and automation systems – handing repetitive work over to machines, and saving time for creativity and decision‑making.

Next, you can easily extend this architecture:

Enterprise AI strategy daily report
Automatic AI knowledge base updates
AI hot topic monitoring system
AI Agent content factory
Dual‑platform (Hugo + WeChat) automated operation

This is the “second productivity system” that indie developers and content creators should invest in during the AI era.