Logo

General Analytics

In this section, you will learn how to analyze the performance of your AI Agent, as analytics play a crucial role in evaluating its effectiveness and identifying areas for improvement to ensure optimal functionality.

Overview

Under this section, you will find basic or general details about your AI Agents. The section includes the following reports:

  1. New Users: This represents the total number of new users who interacted with the AI Agent.
  2. New Sessions: This represents the total number of sessions created by users. For Chat/SMS/WhatsApp the session expires after every 24 Hours. For Voice AI Agent each call is considered a new session.
  3. Messages: Total messages received to the AI Agent.
  4. Response Time: The Average response time took by the AI Agent to respond to user’s query.

Note: The data shown is for a specific AI Agent selected from the Dropdown and for the specific date range selected. If you need the data for a specific channel, then you can select the respective channel from the dropdown provided.

If you want to export these data to a CSV, you can Hover on the respective data and click on 3 dots and click on Inspect and you will get an option to Download data in CSV.

overview

AI Analytics

Under this section, you will find analytics related to your AI Agent. Various functionality analyses are available, helping you identify and differentiate between the most frequently used models.

  1. Classification Model: Any query that is detected or processed by the Intent or Classification model is reflected here.
  2. LLM Queries: Queries answered or processed using LLM prompts are counted in this section.
  3. Cognitive Search: Queries that fall back to the knowledge base are categorized under this section.
  4. AgentM Queries: Queries answered or processed by AgentM are reflected here.
  5. Failed Queries: Queries that are not answered by any of the models are considered failed, and their count is shown in this section.

AI Analytics

Stages

Stages are analytics that act as flags, indicating the last stage, question, or process the caller completed.

For example, if you ask three questions, there can be three stages: Question 1, Question 2, and Question 3. Suppose 10 people called, and 2 hung up after the first question, while the remaining 8 answered all three. For those 8 callers, the stage would be marked as Question 3 because it was the last question they answered.

Stages allow you to track the journey of callers. As the user completes one stage and moves to the next, the stage is updated accordingly.

Stages

Creating Stages within the Workflow

To add stages to your AI Agent, you need to incorporate a JSON API call into your workflow.

For example, if you want to add a stage after a specific process is completed, simply add a JSON API item to the flow. The URL to be used is provided below:

API URL:
https://us.floatbot.ai/voice-api/user-stats?stage=&level=0

The stage Variable should be the name of your Stage. If your Stage name is “Call Successful” then the URL should be as below:

https://us.floatbot.ai/voice-api/user-stats?stage=Call Successful&level=0

Please refer to the screenshot below on how to declare the API in Workflow.

creating stages

× Zoomed Image