From press release
to publication
in minutes.
Redactiefabriek automates the repetitive production tasks that consume your editorial team — so they can focus on real journalism.
A good editor.
An impossible equation.
This is not a bad editor. This is a skilled professional caught in an impossible ratio: more input than output, every single day.
08:27She logs in. 43 press releases. The coffee is still brewing.
08:45First release. 1,200 words of corporate language about an acquisition. She rewrites it into 350 words of readable news. Headline. Subheadings. SEO. Categories. Tags. Publish. That was 50 minutes for one article.
10:30Article five. The pace drops. Headlines become blander. She knows it. But the inbox grows faster than she can empty it.
17:00Eighteen articles done. Twenty-five still waiting. Those will be old news tomorrow. She closes her laptop feeling like she ran all day — and still fell behind.
This is not an exceptional day. This is every day.
Same editor. Same 43 releases.
Different story.
08:27She logs in. Everything is already processed. 43 draft articles ready in WordPress. Written in her tone-of-voice — because the system learned from her previous corrections.
08:35She opens the first draft. Reads it. Adjusts a word in the headline. Publishes. Two minutes.
09:15Twelve articles published. She has the mental space to actually read the thirteenth. It covers the same region as article four. The system already made the connection — a "Read also" link is there.
14:00The afternoon is free for what an editor is actually supposed to do: writing her own story. Depth. Research. The work she chose this profession for.
The system doesn't replace her.
It gives her her profession back.
- 50 to 60 minutes per article, manually
- Quality drops after article five due to fatigue
- Same corrections made over and over
- No time for investigative journalism
- Connections between articles are missed
- SEO metadata often incomplete under time pressure
- Invisible to AI search engines
- Press releases go stale before they are published
- Under 2 minutes review per article
- Consistent quality, 24 hours a day
- Every correction makes the system permanently better
- Afternoons free for in-depth journalism
- "Read also" links placed automatically
- Full SEO, taxonomy and metadata on every article
- AI-findable via 6 technical pillars
- No press release goes to waste
From press release to
publication in 60 seconds.
The system receives content from multiple sources, recognises the format automatically and processes it into a complete, ready-to-publish article.
Email, PDF, Word, Google Drive or API. Every format is detected and processed automatically — including scanned PDFs via OCR.
Promotional language out, journalistic structure in. The AI writes in your editorial voice using your style rules and archive context.
Category, sector, municipality, tags, SEO title and meta description are determined automatically. Related articles are linked.
Score below threshold? Automatic correction round. Only when quality is sufficient does the article appear in WordPress as a draft.
Every layer makes
the next one smarter.
The three modules work independently but reinforce each other. Processing feeds the learning engine. Learning improves findability. Findability brings more readers — and more content to process.
Processing & publication
Every press release, regardless of format or source, becomes a complete draft article in WordPress. Three flows: fully automatic, 1-to-1 placement, or draft for editing. Your choice per source.
Learning & improvement
Every editorial correction is a weighted signal. When a pattern reaches the threshold, it becomes a permanent style rule. The system builds its own style guide — from your actual practice, not from a template.
Visibility & findability
Every article is automatically prepared for Google, Google News and AI search engines like ChatGPT and Perplexity. Geo-tagging, structured data and direct indexing — standard, on every article.
99% of publishers
are invisible to AI.
Millions of people no longer search via Google. They ask ChatGPT, Perplexity or Copilot. These AI search engines work differently — they look for sources that have made themselves machine-readable.
An entrepreneur in Amsterdam opens ChatGPT and asks: "What are the business trends in the Netherlands?" ChatGPT looks for publishers who say: here I am, this is my expertise, and this is what I recently published about exactly this topic.
The result: ChatGPT cites your platform as a source. With name. With link. With context.
At 99% of publishers, nothing happens. Their content exists but is invisible.
Self-hosted.
No vendor lock-in.
Redactiefabriek runs entirely on your own infrastructure. Docker-based, no data leaving your organisation. Delivered as a ready-to-run Docker package.
You manage the updates. You manage the data. Our intelligence runs on your terms.
Technical specifications
"We are not a tech company studying journalism. We are publishers who use technology."
Dennis & Olivier — founders, Redactiefabriek
Let's talk about
your newsroom.
We will show you the system live, using your own content.
Request a demoinfo@fynt.nl · Zuiddijk 412, Zaandam · A product of Fynt Media