Skip to main content
All CollectionsEvaluations
Human vs. AI completed Evaluation
Human vs. AI completed Evaluation

Learn about the difference between a human and an AI completed Evaluation.

Updated over a month ago

Authenticx Evaluations

Evaluations in Authenticx are completed either automatically by AI configured to score answers or by human analysts that listen and complete Evaluations. Both types of Evaluations are sampled into the platform from the full universe of conversations.


Why use AI Completed Evaluations?

AI completed Evaluations allow your team to have a larger surface area on objective questions that you’d like to see automatically answered and scored by the platform. AI completed Evaluations will complete all answers on your Evaluation Forms that are mapped to Editing Smart Predict Answers.

Typically, AI completed Evaluations are used to assess quality scores of agents at a larger scale than can be done by humans.


Accessing AI Completed Evaluations

🔍 Contact your Authenticx representative if you aren’t sure if you have AI completed Evaluations set up.

To find AI completedEvaluations, you can filter conversations that have been completed by the AI by filtering in Conversations.

image.png

Clicking the Blue Checkmark will take you to the specific Evaluation for the conversation. AI evaluated calls are read only, and denote that the answers have been AI completed with the Smart Predict icon.

image.png

Update Answers to AI Completed Questions

If changes need to be made to a read-only, AI-completed Evaluation, the Evaluation must be converted to a human Evaluation.

Why Should Humans Evaluate Calls?

Most organizations use human listening to perform more targeted listening and further enrich insights generated by AI. While AI is able to identify topics, keywords & phrases used on calls, the Eddy Effect (points of customer or patient friction), and more, there are many nuances to conversation that help to have a human ear.

Here are some use cases where human listening is essential:

  • Capturing the nuances of human emotion being discussed in calls

  • Generating a feedback loop

  • Answering additional quality or accuracy questions that may be less straightforward to answer with AI

Accessing Human Completed Evaluations

To find Human completed Evaluations, you can filter conversations that have been completed by humans.

To do this, navigate to Conversations.

Screenshot 2024-10-07 at 11.26.24 AM.png

In the left panel, you’ll notice that you can filter by Evaluation Type. Select Human Evaluated.

image.png

Clicking the green checkmark will take you to the specific Evaluation for the conversation. Human completed Evaluations contain additional nuanced context, including notes:

Additionally, when humans change the Smart Predicted answer, the icon disappears. This indicates that a change has been made to the AI's predicted answer, while still retaining the original response.

ℹ️ Learn how to complete an Evaluation.


The Harmony of AI and Humans

Together, AI and Human completed Evaluations offer a multitude of benefits, including but certainly not limited to:

  • Rich context that adds structure behind challenging, human-oriented questions

  • Creation of a feedback loop through human validated AI scoring

  • Clips with examples of specific insights from conversations

  • Coaching feedback that is scored for quality and accuracy

Did this answer your question?