E1027: Tryggvi Rafn rejoins the show to screenshare (on YouTube) the AI-powered SEO system he built from scratch over four months.
This is not another "AI writes blog posts" workflow. The system is designed to discover, prioritize, and build revenue-driving pages automatically, using a combination of data pipelines, keyword clustering, and agent-based workflows.
We walk through how it works end-to-end, including how it finds opportunities, creates content, and routes everything through a structured review process.
What we cover:
- How the system builds a "keyword universe" from multiple data sources
- Using Google Search Console, scraped content, forums, and product data together
- Turning website content into a vector database to better understand topics
- Automatically generating and scoring new keyword opportunities
- Clustering keywords into topics and mapping them to the right pages
- Why most tools focus too much on blog content (and what to do instead)
- Building commercial pages, landing pages, and product-driven SEO content
- How the system identifies missing pages and creates them
- The role of intent analysis and why it needs to be tracked over time
- How "striking distance" and information gain are used to improve rankings
- The content pipeline: research → brief → draft → review → publish
- Why human review is still a critical step in the workflow
- How agents are used for research, analysis, and execution
- The difference between stateful agents and task-based agents
- How the team iterates on content quality and improves outputs over time
- What actually makes AI-generated content useful instead of generic
- The biggest challenges in building a system like this (and what broke along the way)
- How much time this saves compared to manual SEO workflows
- Real-world results, including ranking for competitive keywords and driving high-ticket sales
We also get into the technical side:
- Building data pipelines that don't break at scale
- Error handling and workflow reliability
- Managing handoffs between different agents
- Designing systems that continuously improve with feedback
This episode is especially useful if you:
- Run SEO for an e-commerce or service business
- Want to move beyond manual keyword research and content planning
- Are thinking about building internal tools using AI
- Care about ranking pages that actually drive revenue, not just traffic
⭐️ Tryggvi Rafn on 𝕏 - https://x.com/ecomtryggvi
⭐️ Tryggvi Rafn on LinkedIn - https://www.linkedin.com/in/tryggvi-rafn/
⭐️ Tryggvi Rafn on YouTube - https://www.youtube.com/@Ecomtryggvi
⭐️ Tryggvi Rafn's agency - https://www.nordicamarketing.com/
🚀 Edward's SEO Articles - https://edwardsturm.com/articles/search-engine-optimization/
💎 Compact Keywords - My SEO Course - Get paying customers through SEO - Clear step-by-step video breakdowns - SEO templates to be copied and adapted for your products and services: https://compactkeywords.com/
00:00 Auto SEO Overview
01:20 Why Page Types Matter
02:06 Brand DNA Brain
03:05 Keyword Universe Pipeline
06:38 Clustering Explained
07:47 Results and Human Review
08:46 Command Center Workflows
11:02 Intent and Commercial Focus
11:57 Images and Alt Text Automation
14:29 Project Brain Knowledge Base
17:46 Information Gain Briefs
19:45 Backlog to Review Workflow
24:54 Writing Quality Controls
26:36 Building Lessons and Roadblocks
29:52 Five Keyword Sources
32:54 Impact on Rankings
33:47 Wrap Up and Next Features
The Edward Show. Your daily search engine optimization podcast: https://edwardsturm.com/the-edward-show/
#searchengineoptimization #seo #programmaticseo #growthhacking