Our editorial team,
augmented by AI.
The context
DCKAY Agency is also a newsroom. Coralie writes for six French magazines: Confidentiel, Public, Magic Maman, Le Nouveau Détective, Icon Life and others. Steady volume, hard deadlines, non-negotiable press quality. And on our side, we preach automation for SMBs. We had to apply it to ourselves.
The problem
Topic research, sourcing, angle structuring: every one of these steps ate hours every week on low-value work. Coralie was spending a third of her time preparing pieces before she even started writing. We wanted to give that time back to writing and editorial direction, what she does better than anyone.
What we built: an AI-driven editorial calendar
We built a shared editorial calendar in Notion, fed directly by AI. In practice, we talk to the agent in natural language: « Add the TRIX magazine to the editorial calendar, due to press on February 20 », or « I've finished delivering TRIX, move it to delivered status ». The agent understands, updates the Notion database, and notifies the right people. The calendar is followed and updated by every contributor: writers, designers, proofreaders. It serves as the operations base before sending to print. No more Excel files drifting out of sync, no more email chains asking who's doing what.
What we built: automated AFP news monitoring
The AI agent shares AFP news with us every morning, via email and Telegram. We reply with a topic, and the n8n workflow behind it automatically updates the magazine table of contents in Notion and assigns the topic to the relevant writer, who gets a notification. Editorial selection stays 100% human, but the plumbing around it disappears. We save an hour where it used to take three.
What we learned
Three things. First: AI does not replace a writer, it replaces the tasks around the writer. Second: quality depends on the quality of the brief you give the agent, so the real craft becomes brief design. Third: an automation workflow is built in short iterations, not in one big project. We shipped a first version in two weeks and refined it over six.
What this means for you
We apply the exact same method to your company. We look at what repeats, identify what can be delegated to a machine without losing quality, build the workflow, iterate. You recover time where your team was losing it. And you keep control of the decisions that matter.