Conversation Completeness

Conversation Completeness is a multi-turn metric to determine if a conversation is complete.

Overview

The conversational completeness metric is a multi-turn metric that uses LLM-as-a-judge to evaluate whether your chatbot satisfies the user’s requirements at each turn throughout the conversation.

Required Parameters

These are the parameters you must supply in your test case to run evaluations for conversation completeness metric:

turnslist of Turnrequired

A list of Turns as exchanges between user and assistant.

Parameters of `Turn`:

roleuser | assistantrequired

The role of the person speaking, it's either user or assistant

contentstringrequired

The content provided by the role for the turn

How Is It Calculated?

The conversation completeness metric first extracts distinct user intentions from all turns using an LLM, then uses the same LLM to check if the corresponding assistant turns have satisfied those intentions.

\text{Conversation Completeness} = \frac{\text{Number of Satisfied User Intentions in Conversation}}{\text{Total Number of User Intentions in Conversation}}

The final score is the proportion of satisfied user intentions found in the conversation.

Create Locally

You can create the ConversationCompletenessMetric in deepeval as follows:

from deepeval.metrics import ConversationCompletenessMetric

metric = ConversationCompletenessMetric()

Here's a list of parameters you can configure when creating a ConversationCompletenessMetric:

thresholdnumberdefault: 0.5

A float to represent the minimum passing threshold.

modelstring | Objectdefault: gpt-4.1

A string specifying which of OpenAI's GPT models to use OR any custom LLM model of type DeepEvalBaseLLM.

include_reasonbooleandefault: true

A boolean to enable the inclusion a reason for its evaluation score.

async_modebooleandefault: true

A boolean to enable concurrent execution within the measure() method.

strict_modebooleandefault: false

A boolean to enforce a binary metric score: 0 for perfection, 1 otherwise.

verbose_modebooleandefault: false

A boolean to print the intermediate steps used to calculate the metric score.

Create Remotely

For users not using deepeval python, or want to run evals remotely on Confident AI, you can use the conversation completeness metric by adding it to a single-turn metric collection. This will allow you to use conversation completeness metric for:

Multi-turn E2E testing
Online and offline evals for traces and spans