> ## Documentation Index
> Fetch the complete documentation index at: https://docs.lorikeetcx.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# How hard the work is

> How much of your queue an answer alone could handle.

The **Answerable from knowledge alone** tile shows the share of tickets an answer could have resolved, without looking anything up or changing anything.

## How it's measured

An AI model rates each ticket from its conversation into one of three tiers. The ratings are estimates, not a record of what your team did.

| Tier | What it covers |
| - | - |
| **Answer from knowledge** | An answer is enough |
| **Look up or one change** | Looking something up, or one change in one system |
| **Several steps or a person** | Several systems, or a person, such as a complaint or a fraud check |

Auto-replies, spam and a bare "thanks" are left out. If most tickets in your file have only the customer's first message, the model has less to go on, and the report says "up to" for the hardest tier.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.