We’re continuing to invest in the Conversation Details view, as another part of our “Connect the Dots” efforts. Here are the latest updates:

When you create a clip from the audio player, if you have the transcript open, you’ll see the transcript text highlighting in real-time as well.

When playing a call, the transcript text will highlight, so you can read along as the audio plays.

Additionally, Eddy highlights are back in the transcript!

We have made significant investments in KBo this quarter, and are delighted to share a few enhancements with you here!

KBo’s sample analysis is stronger if conversations have more in common. For example, looking at calls from a particular brand from patients with an Eddy will result in a more useful analysis than just a set of very diverse conversations from the past week.

Now, KBo will encourage users to apply filters and narrow down their search if it’s very broad when they request an analysis.

Once you’ve narrowed down and are ready to analyze, KBo will now ask if you have a specific question in mind. If you do, KBo will do the transcript analysis while trying to answer that question as much as possible.

This is huge for allowing users to get an even more targeted analysis of the calls they care about.

Examples of what users could ask, depending on their goals:
​
​Customer Understanding:
​“What reasons did customers give for stopping therapy on these calls?”

Agent Performance:
​“How well did the agent explain the patient assistance program?”

Process Adherence & Compliance:
​“Did the agents consistently verify the caller’s identity?”

Emotional Tone:
​“In what ways did callers express frustration or confusion during these conversations?”

Outcome Tracking:
​“What actions did the agents commit to at the end of these calls?”

Now, after KBo analyzes a sample of conversations, users can ask open-ended questions about the analysis, and KBo will reference the content of those analyzed conversations to provide deeper insights.

We have run into a great problem: customers are making SO many custom dashboards! To help with organizing all of them, we’ve done two things:

We have a fancy new way for users to find and select agents for “group by” on the Model Results report in Signals! Whereas before, users needed to remember the names of agents they wanted to select, now, users can peruse a full list of agents, search for agents by name, and make multiple selections. This makes it easier than ever to find and choose agents for grouping! Expect a similar update to the agents filter in Signals and Conversation Explorer, coming soon.

Instead of showing Spotlights fully expanded by default on every Signals graph, users can now click “Analyze with KBo” when they are interested in getting more information about the data on the graph. This helps visually “declutter” your Signals graphs, while giving you on-demand analysis from the Authenticx AI Assistant, KBo.

Now, users can click a P2 transcript highlight (for Eddy or Adverse Events) to go directly to that part of the conversation in the transcription tab! This makes it easier than ever to find the relevant part of the conversation. From there, you can quickly create clips, turn those into a montage, or include that part of the transcript in your analysis. Additionally, those P2 highlights are labeled in the transcription.

Customers with access to KBo and model builder models will now be able to generate a full, deep analysis of results from models created in Model Builder — in a matter of seconds, using KBo.

This gives users immediate insight into why model outcomes occur across all conversations, without manually reviewing model reasoning within individual conversations, one-by-one. That’s right: this unlocks high-quality, AI-powered analysis, at scale, with the click of a button.

Users can go deeper on a single result, or on model results in general.

Both analyses leverage model reasoning to produce richer insights, helping users dig into outcomes more efficiently.

For quality models, this allows users to ask questions like “when the Agent Consideration result is Does Not Meet, what’s generally going wrong?”

For insights models, users can use this option to learn more about a given result of the Social Determinants of Health model, for example.

If your team does not have access to KBo, clicking “Go Deeper” will give you the opportunity to get in touch with the Authenticx team to learn more.

Last week, we introduced our first Smart Banners in the platform in Conversation Explorer! These banners help users understand “why am I seeing what I’m seeing here?”

With this week’s release, in addition to Smart Banners in Conversation Explorer, users will now see Smart Banners in Conversation Details. These Smart Banners will highlight why this specific conversation is relevant for you, based on the filters you applied in your Conversation Explorer search.