Auto-Suggest Knowledge Base Articles
Nora reads the ticket and suggests relevant help articles to resolve it faster
Many support issues are answered in your help center. This automation reads incoming tickets and suggests matching knowledge base articles to the agent, avoiding duplicate answers.
Support teams have knowledge bases, but agents don't always use them. They re-answer questions that are already documented, wasting time and causing inconsistency. This automation monitors incoming Freshdesk tickets. When a new ticket arrives, Nora reads the subject and description, extracts key terms (e.g., 'password reset,' 'billing,' 'export'), and searches your Freshdesk knowledge base for matching articles. She then surfaces the top 2-3 matching articles to the agent in a sidebar, ranked by relevance. The agent can click 'Link this article' to attach it to the ticket, or 'Ignore' if it's not relevant. Over time, Nora learns which articles solve which types of tickets, and improves her suggestions. Example: A customer writes: 'How do I reset my password?' Nora reads this, finds your 'Password Reset' knowledge article, and suggests it to the agent. The agent clicks 'Link,' and the article is automatically sent to the customer. Customer self-serves, no agent time spent. Ticket closes in 60 seconds.
How it works
- Nora reads incoming Freshdesk ticket subject and body
- Extracts key terms and intent
- Searches knowledge base for matching articles
- Scores and ranks articles by relevance
- Displays top 2-3 suggestions to the agent
- Logs which articles are used vs. ignored
Frequently asked questions
What does "Auto-Suggest Knowledge Base Articles" do?
Many support issues are answered in your help center. This automation reads incoming tickets and suggests matching knowledge base articles to the agent, avoiding duplicate answers.
How long does setup take?
Under 5 minutes. Connect your tools and Nora handles the rest.
Which tools does this use?
Uses Freshdesk, run by your Customer Support.
How much time does this save?
Saves approximately 2 hours/week, running real-time (on new ticket).