Freddy Copilot’s Reply Suggester is a powerful feature designed to help agents resolve tickets faster and more efficiently. By suggesting the best responses sourced from your organization’s knowledge base, Reply Suggester reduces resolution times, improves customer satisfaction, and enhances agent productivity.

This article contains

About Reply Suggester

When agents receive a ticket, they often need to,

  • Search for relevant solution articles.
  • Extract the right information from these articles.
  • Craft a professional and accurate response.

This process is time-consuming and can lead to delays in ticket resolution.

Reply Suggester automates this process by,

  • Suggesting the best response to incoming ticket queries via the email channel.
  • Sourcing relevant content from your KBase, including public, private, and bot articles.
  • Citing source articles so agents can validate the suggested responses.

Key considerations and best practices

  1. Reply Suggester works only for tickets created via the Email channel or through Portals.
  2. It suggests responses only for the first customer reply in a ticket.
  3. Requires Solution Article Suggester to be enabled to view the sources in the suggested reply.
  4. Maintain a well-organized and up-to-date knowledge base of common customer issues.

Business use cases for Reply Suggester

Use case 1: Improve resolution times for high-volume L1 queries

Scenario: An e-commerce company receives hundreds of daily queries about order status, returns, and product information.

Solution:

  1. Enable Reply Suggester in the Freshdesk admin panel.
  2. Agents receive auto-suggested responses for common queries like “Where is my order?” or “How do I return a product?”
  3. Agents validate the response using cited solution articles and send it to the customer.

Impact: Resolution times drop, and customer satisfaction improves.

Use case 2: Enhance agent productivity and reduce training time

Scenario: A SaaS company onboarded ten new agents who find it challenging to create accurate responses to technical queries.

Solution:

  1. Enable Reply Suggester and ensure the KBase is updated with technical solution articles.
  2. New agents use auto-suggested responses for queries like “How do I reset my password?” or “Why is my dashboard not loading?”
  3. Agents edit responses as needed and provide feedback to improve accuracy.

Impact: New agents achieve productivity faster, and training costs are reduced.

Use case 3: Ensure consistent and accurate responses

Scenario: A telecom company faces complaints about inconsistent responses to billing and plan-related queries.

Solution:

  1. Enable Reply Suggester and update the KBase with billing and plan information.
  2. Agents use auto-suggested responses for queries like “Why was I overcharged?” or “What are my current plan details?”
  3. Agents validate responses using cited articles and send them to customers.

Impact: Response accuracy improves, and customer complaints drop significantly.

Enable Reply Suggester

As an Admin, you can enable the feature by following the steps below,

  1. Navigate to Admin > Freddy > AI Copilot. If you are using a Freshdesk Omni (formerly Customer Service Suite - CSS) account, then go to Admin settings > Freddy AI and Self Service > Freddy for Ticketing > AI Copilot.
  2. Toggle the Reply suggester switch to activate the feature.

Use Reply Suggester

  1. Open a ticket created via the email channel or through support portals.
  2. A red dot appears on the Reply button if a suggested response is available. Click Reply to open the editor.
  3. The AI-generated response automatically appears as ghost text(grayed out) in the reply field.
  4. If an agent starts composing a reply and needs AI assistance, while typing in the reply editor, click the Suggest reply option under the Freddy icon in the toolbar.
  5. Based on the ticket content, Freddy AI will generate a response as a ghost text, replacing any existing draft.
  6. Accept the response by pressing Tab or Enter, or edit it as needed.

  7. A banner displays the number of solution articles used to generate the response.

  8. Click the banner to view the articles in the right pane (requires Solution Article Suggester to be enabled).

  9. Agents can rate the accuracy of the suggested response by clicking on the thumbs-up/thumbs-down option.