Company Case Study

Development of a News Monitoring System with AI-Powered Rewriting

Media

A case study of building an admin platform for content teams: news monitoring, automated AI rewriting, and WordPress publishing.

News Monitoring and AI Rewriting System

About the Project

Media / Content Platforms / AI-Powered Tools • Web / Admin Panel

News Grabber is an internal product developed as a unified admin system for content teams and news websites.

The platform covers the entire newsroom workflow: from search query to WordPress draft, with a strong focus on automation and quality control.

AI RewritingNews MonitoringWordPress Integration

Project Context

Content teams and news websites face the same operational challenges every day:

News has to be sourced manually across multiple channels
Articles cannot simply be copied — unique content is required
Rewriting takes a significant amount of editorial time and effort
Publishing to CMS platforms like WordPress requires switching between tools
Scaling content operations across multiple topics is difficult

News Grabber was designed as a single admin system covering the full workflow: from topic query to a ready-to-publish draft in WordPress.

Business Goals

Monitor news by topic

Search via GNews using LLM-generated keywords

Save editorial time

Automatically rewrite articles with AI

Produce unique content

Custom rewrite prompts with style and fact control

Scale topic coverage

Mass Search and dashboard management

Avoid missing new stories

A “new articles” counter for every search query

Publish quickly to WordPress

Export drafts to WordPress in one click

Maintain source transparency

Automatically attach a link to the original article

How the Product Works (User Flow)

1. Topic-Based News Search

  1. 1The user enters a topic (for example, “Donald Trump” or “Climate change”)
  2. 2

    The system:

    • • generates keywords via LLM
    • • searches news via GNews
    • • stores the discovered articles
    • • automatically rewrites each article via LLM
  3. 3The user receives a page with article cards

2. Mass Search

  • Run searches across multiple topics at once
  • Queries are processed sequentially, with visible progress tracking
  • Results for each query are saved and available in history

3. History and Dashboard

All search queries are stored. For each query, the system shows:

  • last update date
  • total number of articles
  • number of new articles

The dashboard allows users to:

  • bulk refresh selected queries
  • sort topics by freshness
  • quickly identify the most active topics

4. Article Workflow

Each article includes:

  • the original text with a source link
  • the rewritten version
  • generation and publishing status
  • a WYSIWYG editor for manual editing

Available actions:

Regenerate rewrite
Save edits
Send to WordPress
Update draft

5. Bulk Rewriting

  • Select up to 10 articles at a time
  • Choose a prompt from the library
  • Launch bulk rewriting with a single command

6. Prompt Management

A dedicated settings section includes:

  • a prompt for keyword generation
  • a rewrite prompt library with model selection, temperature settings, and default prompt configuration
  • a testing area for trying rewrites without saving

This gives the team control over style, tone, and output quality instead of relying on a black box.

The Role of AI in the Product

AI is not used as a standalone feature, but as a core part of the product logic:

Generating keywords for a topic
Rewriting articles while preserving facts
Maintaining a neutral news-style tone
Regenerating content based on updated rules

Important: AI does not replace the editor — it removes repetitive work, while leaving final editorial decisions to a human.

Solution Architecture

Backend

  • Node.js + TypeScript
  • Express
  • MongoDB
  • OpenAI API integration
  • GNews integration
  • WordPress REST API integration

Frontend

  • React
  • Vite
  • Tailwind CSS
  • React Query
  • Zustand

Infrastructure

  • Docker Compose
  • Nginx
  • Separate frontend / backend / database services

Technology Stack

Frontend

ReactViteTailwind CSSReact QueryZustand

Backend

Node.jsTypeScriptExpress

Database

MongoDB

AI / Integrations

OpenAI APIGNews APIWordPress REST API

DevOps

Docker ComposeNginx

Key Project Challenges

Technical

  • Handling bulk processing without overloading APIs
  • Asynchronous rewrite generation
  • Managing article and search statuses
  • Republishing without creating duplicate drafts in WordPress

Product

  • Preserving facts and a neutral editorial tone
  • Maintaining content quality under automation
  • Keeping the editor experience efficient without overloading the UI

Solutions

  • Queues and statuses for searches and rewrites
  • A prompt library instead of a single “universal” prompt
  • Clear separation between original content and rewritten content
  • Each article is linked to a single WordPress draft

Results

End-to-end news workflow

In a single interface

Significant time savings

For content preparation

Scalable news monitoring

Across dozens of topics simultaneously

Higher content quality

And greater uniqueness of published articles

A practical tool

For editorial teams and content operations

What the Team Is Proud Of

A systems approach: from monitoring to publishing
Controlled use of AI, not “magic”
Thoughtful UX for daily editorial work
Scalability without increasing manual workload

Key Takeaway

News Grabber is more than just a news rewriting tool.

It is a workflow platform for editorial teams that turns the news stream into a structured, scalable, and controllable process — from query to WordPress draft.

Want to discuss your project?

Get in touch, and we’ll show you how we can help bring your idea to life