Welcome to day 23 of our countdown to the New Year! At Asana, keeping our Help Center articles accurate and useful is a top priority. To make this process seamless, we’ve built a system using AI and rules. Here’s a quick breakdown of how it works.
We’ve created an Asana project designed to streamline article reviews. Each section of this project corresponds to a theme or category in our Help Center, making it easy to organize and review content effectively.
Create rules for each section
Each section in the project has its own rules to automate the review process. The rule is set up so that when a task is moved into a specific section (e.g., “Getting Started”), a subtask is automatically created, including a response generated by Asana AI.
You need to provide AI with specific instructions depending on what you want it to review or suggest. In our case, the AI is instructed to review the attached article and check for references to specific dates or timeframes that might be irrelevant now. We provide the AI with attachments of similar articles as guardrails to ensure consistency, so its suggestions align with our content standards. This setup ensures every article is thoroughly and efficiently reviewed.
How we review articles
- Create a task for your content
If an article hasn’t been reviewed in six months, a task is created with a PDF of the article attached and added to the “Intake” section.
- Move your content to the relevant section
Based on the article’s topic, we then move our task into the appropriate section of our Asana project.
- AI Suggestions
Asana AI creates a subtask with detailed recommendations based on the rule for that section.
This system keeps our content fresh and saves time by providing targeted suggestions for updates and allowing collaboration between reviewers and Asana AI!
