Analyze your conversations faster with AutoLearn

Carl Schultze

November 3, 2022

Tethr’s newest machine learning capabilities enable you to build structured insights with incredible ease, speed, and accuracy.

Speech analytics and conversation intelligence platforms help companies understand their customer experience.

Standard conversation intelligence and speech analytics platforms all rely on humans to program and write code in order to detect events in your conversations. This time-consuming process often requires special training to learn how to write the code conversation intelligence platforms accept.

Many conversation intelligence platforms, including Tethr, call the detected key moments in conversations “categories.”  Once you’ve programmed the platform to look for specific phrases and its variations, it will sort through all your conversations and find those that include that category.  From there, you can use that structured category to filter interactions, build reports, trigger alerts, and many other possibilities.

Custom categories

While Tethr includes some standard categories already, we know that what specific terms you might want to listen for can depend on your unique business. For example, if you produce consumer goods, you would likely want to monitor conversations for your specific product names so you could delve into what customers say about each one.

AutoLearn eliminates the need for manually programming categories

At Tethr, we’ve  entered a new era in category building. We’re excited to introduce our latest machine learning capability: AutoLearn.

AutoLearn leverages AI-powered learning to make the process of creating new categories simple and fast.

With AutoLearn, you get:

  • Incredible acceleration in building structured insights
  • No more manual programming
  • Significantly reduced time in training
  • Build new categories with incredible accuracy and ease
Download The End of Empathy Ebook

How Tethr makes building structured insights easier

We call the process of  combing through all your conversations and teaching your platform to recognize these phrases “scripting.” Using older methods, it can be time consuming and require you to think of all the variations in how people might describe something.

With Tethr’s latest update, that scripting process is now enabled with AI-powered learning to make the process of creating new categories simple and fast.

Anyone can create new categories with our new feature, AutoLearn, without needing training, in three easy steps. Here’s how it works:

  1. Create a new category

Start by entering the must-have words that you want to capture. For example, if you want to see how often someone calls to ask for a refund, you could enter the word “refund” first. We recommend only including the must-have words so that Tethr can capture all the nuances in how people express themselves.

  1. Have AI search and capture phrases

Once you’ve added your “must-have” words to the list, you would select “Search and browse,” and Tethr’s AutoLearn feature goes to work, finding examples of potentially related phrases from your conversations. Add those that match the situation you’re trying to measure.

In our example, “refund” might include phrases like “need a refund,” or “have a refund” but you might eliminate a phrase that isn’t a direct match, like “refundable deposit,” which occurs when people discuss a separate issue.

  1. Validate findings

After Tethr AutoLearns the phrases, you can check your work. Using our “Validate” feature, Tethr shows you which phrases appear like a match to the key phrase that you were searching for. If any false positives show up, you can remove them from consideration.

The easiest conversation intelligence platform to use

We strive to make Tethr easy to use. We’re constantly working to improve our platform so that it’s simple and effective. Want to see for yourself what kind of analytics you can get out of your conversations? Start a free trial.

Download Recession-Proof Ebook
Jump to:

Most popular articles