Speech & text analytics

Decode what’s happening in your CX conversations, across all channels.

Speech recognition, transcription, NLU (Natural Language Understanding) and keyword spotting in 30+ languages. SuccessKPI’s AI/ML engines analyze millions of transactions across all channels at scale to surface trends and patterns.


Turbocharge your data to discover “what you don’t know.” And empower yourself to take action.


Speech and text analytics begins with realtime transcription and Natural Language Understanding in 30+ languages

SuccessKPI detects sentiment, themes and topics from realtime media streams, call recordings or text transcripts immediately following the conversation.

The platform itself works in 10 different languages— while it can transcribe, understand, and analyze voice and text streams in up to 30+ languages.

Start with the out-of-the-box Playbook library. Then, customize the topics and themes, and set up real time automated actions on your call center conversations.

Our AI powered speech analytics for Amazon Connect, Genesys Cloud, UJET, and other top CCaaS platforms allows you to see inside your contact center conversations.



Spot keywords, correlate themes, and understand sentiment at the conversation, brand and entity level.

SuccessKPI comes pre-loaded with dozens of common call center topics, like politeness, mini-miranda, and ownership. Or, build any topics you need from as many keywords that occur in your environment to make your analysis complete.

Topics can be read at the agent, customer or dual channel at the same time. Build complex themes by combining and weighting topics. Our ML engines will find new words and keyphrases to correlate with your logic, and they automatically identify brands and locations in realtime.


A full view of speech and text analytics involves decoding the language at multiple levels


Topics are collections of keywords and phrases relevant to your customer experience strategy. Topics can represent concepts, such as “politeness” or “ownership” — or key business elements like product names and store locations. Within topics, you assign all the keywords that occur in conversations, related to the topics.


Themes are score-weighted collections of Topics, and you build them into more complex concepts. For example, a theme like Positive Agent Behavior could be built from the Topics “Politeness,” “Proactive Language,” and “Agent Compliments.”

Custom phrases

In many business operations, keywords are not easily understood by generic speech recognition and NLU engines. Use custom phrases to ensure that specialty words unique to your business are properly documented by SuccessKPI.


Sentiment can be positive, negative or neutral and can be based on the agent, the customer or the entire channel.


Redactions represent keywords that you don’t want transcribed or stored for business purposes— for instance PII (Personally Identifiable Information) such as credit card, social security or telephone numbers.

Speech and text processing are the beginning…
Take them to work with automation and next-level analytics.

Unlock the power of speech and text analytics with Playbooks and data discovery in our enterprise BI Layer


Playbooks for realtime action

Drive automated outcomes based on business objectives. Playbooks have three primary components:


Topics detected, sentiment, a combination of topics and themes, and metrics.


Logic allows you to reference ALL, ANY, or SOME of the conditions.


Send emails or SMS messages, hit REST APIs (like log a case in Salesforce), invoke a lambda function, or send a call to the evaluation workspace.

How does it work? A sample scenario to analyze and improve churn:

  1. Set up a Playbook to identify any calls with negative sentiment on premium customers where the cancellation topic is detected.
  2. Set up a Playbook action to push these calls to the evaluation workspace and log a case in your support system.
  3. Analyze the common attributes and causes of cancellation, then create a new Playbook to escalate with an email alert to a supervisor in realtime for the most critical cases.

Data discovery for deeper understanding

The enterprise data warehouse and BI layer are a deep resource of information.

Every conversation is transcribed, processed and stored by our realtime AI engines in the enterprise data warehouse. Perform advanced data discovery and integrate the speech and text analytics data with hundreds of other data sources.

All of your contact center metrics including IVR outcomes, chat conversations, Average Handle Time, Agent Performance data, call duration, After Call Work, are imported, curated, and blended with the speech and text analytics layer.

Integrations allow you to connect data from CRM, marketing, campaign, billing, collections, case management, and other operational and business support systems that contain data that you need to complete your 360° view.

An industry-first total conversation view

Bring your speech and text data to life with new insights, and understand the customer experience on a new level.