Challenge

Anyone working in publishing needs to know what is happening in the industry — which stories are going viral and, more importantly, which ones are actually relevant to their own work. At this client, that screening was spread across several people from several teams: each of them spent 30 to 60 minutes per week researching industry news, reading it, assessing it, and passing it on to the colleagues responsible.

The outcome was what you would expect from distributed manual work: the effort added up across the organization, and the screening depended on individual people and the time they happened to have available.

Approach

In a four-week project, we automated the entire pipeline — with n8n as the workflow orchestrator and ChatGPT on Azure as the analysis engine.

The agent pipeline has since taken over three tasks that people used to do:

  1. Research and summarization: Viral and relevant industry news is found, summarized, and content-analyzed automatically.
  2. Relevance assessment and routing: For every article, the agents assess which departments it is relevant to — and decide, based on the known responsibilities, where it belongs.
  3. Distribution: Depending on the topic, the relevant teams are notified automatically.

Outcome

Manual screening has been eliminated entirely. Instead of researching, reading, and distributing themselves, the people involved now only skim the distilled results — and take the finally relevant items straight into their meetings for further processing.

What used to be 30 to 60 minutes of screening work per person per week — across several people and teams — has become a quick look at a ready-made distillate.