Cresta Launches Next-Gen Insights with New Real-Time Trends

On July 29, Cresta CEO Ping Wu announced the next generation of Cresta Insights, expanding the AI platform’s existing conversational intelligence capabilities to include new Real-Time Trends while deepening existing AI Analyst and Topic Discovery insights.

Cresta Insights already helps businesses discover “unknown unknowns”, analyzing 100% of omnichannel customer conversations and answering the questions most companies never would have known to ask. 

These updates mean:

• Cresta’s AI Analyst no longer just answers user questions, but autonomously plans and conducts conversational intelligence research, generates a variety of interaction reports, and provides next-best-action suggestions to reach desired outcomes
• Topic Discovery, which already automatically grouped conversation by reason via LLM clustering, now offers a nuanced multi-layered view of common topics, creating consistent taxonomy while ensuring human-in-the-loop oversight
• New Real-Time Trends identifies and alerts teams about real-time customer conversation anomalies, creating an auto-generated trend headline and leveraging AI Analyst to identify potential root causes

Per Wu, Cresta’s goal here is to create “systems that can maintain a reliable understanding of customers’ needs, detect emerging changes without being explicitly told what to look for, and perform more of the analytical work required to turn those signals into data-driven decisions. We’re one step closer to our vision: one proactive customer intelligence system that doesn't wait for someone to ask the right question.”

Cresta AI Analyst

Cresta launched AI Analyst in early 2025 as a smart query tool that let users ask questions about their business in plain English. This announcement turns AI Analyst into a deep, autonomous research agent with deep planning and multi-step reasoning capabilities. In short: AI Analyst both finds problems and tells you how to solve them. 

Users no longer lose time to manually filtering data, deciding on report types, and drafting them. Now, AI Analyst chooses the best type of report to run based on the user question, finds the relevant data/conversations, analyzes them, then creates multi-modal comparisons, charts, and briefs based on the results.

What’s unique here is that, in addition to raw voice transcripts, AI Analyst draws on Crestas’s additional conversation data for greater context. (Think key conversation moments, topics, and categories, agent/customer behaviors, and objections set in Cresta Opera.)

The new AI Analyst provides greater accuracy, deeper insights, and clear next steps: businesses understand what’s happening, why, and what to do about it.

Topic Discovery 

Topic Discovery historically used LLM clustering to categorize conversation topics. Cresta’s latest announcement essentially makes it much easier for businesses to keep that automatically generated taxonomy accurate and in line with strategies over time. 

Now, Topic Discovery generates categories and up to three layers of subcategories, instantly surfacing the natural structure that already exists in your customer conversations. From there, the human-in-the-loop takes over, as CX teams can review, merge, rename, and rearrange proposed categories to match internal terminology.

Once that taxonomy is clearly set, it serves as a single source of truth for the reasons behind customer calls. Every new conversation is automatically tagged (no manual keywording required). These established categories carry across Cresta in dashboards, AI Analyst queries, automated workflows, and even live agent guidance.

Real-Time Trends

Real-time trends is the entirely new feature of the Cresta Insights Suite. While historical conversational intelligence and better topic categorization help, they can’t fix the real-time issues or predict reactions based on live customer activity.

Its stated specialty is unknown-unknown coverage. That means catching keywords, phrases, and events that don't yet exist in any taxonomy (outages, regional disruptions, a newly-broken product feature, etc.)  

Powered by AI Analyst, every new trend arrives with an auto-generated headline and root-cause explanation (seen above.) Subscribers get proactive email alerts the moment a topic starts to spike.

The CX Impact 

Cresta’s announcement echoes the current focus on the shift to agentic AI that strategizes and acts instead of simply reports. Cresta's specific angle is proactivity and portability: intelligence that surfaces itself, identifies its own root cause, and immediately triggers relevant action inside the same platform.

Taken together, these tools all make the argument that the burden of knowing what to look for should not fall on the user. Cresta’s vision is one of a proactive, agentic conversation intelligence system that discovers what customers are calling about, why, and what to do next without anyone asking the right question first. While these three updates don’t provide an entirely autonomous system, they represent a coherent and credible step toward it.

Expect this tighter coupling between Insights, AI Agents, and Agent Assist to be a centerpiece at the Cresta WAVE conference in Nashville this December.