Cross-Mining
Speech to text : Automatic transcription of your customer calls.
Cross-Mining Converts the audio from your conversations into time-stamped text that can be used for conversation analysis, " Quality Monitoring ," and generative AI. Whether recorded or live, via SaaS or on-premises, using engines from CrossCX and its partners specializing in speech recognition.
Recorded for analysis, live for immediate response.
Automatic conversation transcription is available in two distinct workflows. You can choose the one that best suits your needs: analyzing large volumes of data after the fact, or having the text available during the conversation.
Batch: Offline Transcription
Recordings are transcribed after the call, in batches. This method is best suited for large volumes and in-depth analysis.
- Transcription of call recordings retrieved by the connectors
- Data for conversation analysis and quality assessments
- Text archived along with its metadata for search and reporting purposes
Live: Live Transcript
The text is generated as the conversation progresses, for situations that can't wait until the end of the call.
- Signal Detection During the Exchange
- Real-time alerts sent to the right people
- The same text can then be reused for analysis and quality assurance
From audio to usable text, in four steps.
Transcription is not an end in itself: it is the raw material for the entire analysis process at Cross-Mining.
Record audio
Recordings are imported from your telephony and customer relationship management systems throughthe interception and centralization of conversations.
Speech Recognition
The speech recognition engine converts the audio signal into words in more than 30 languages.
Organizing the Text
Each transcript is time-stamped, tagged with metadata (agent, queue, campaign, date), and archived.
Analyze and Take Action
The text is used forsentiment analysis, categorization, summarization, and translation.
A reliable document, three tools for your teams.
Analyze all conversations
Topics, reasons for contact, trends, and warning signs—based on all calls, not just a few monitored calls.
Conversation analysisAssess quality more quickly
Read instead of listening again, find the passage that justifies a grade, and automate part of the grading process.
AI-Driven Automated Quality ManagementMeasuring the Fluidity of Exchanges
Rate of speech, overlapping, and silences at key moments: the paraverbal indicators that complement words.
The Three Drivers of AnalysisThe engine tailored to your needs, whether in the cloud or on-premises.
The world of AI is evolving very quickly. CrossCX 's analytics connectors allow you to choose—and then scale—the transcription engine that best suits your languages, volumes, and security requirements.
View all analytics connectors and connectors to generative AI.
"We can sample calls and easily apply our listening checklists to them, and the Speech Analytics module performs very well."
RATP
Speech to text : Your questions.
What is the " speech to text "?
speech to text, or speech recognition, automatically converts speech into text. In contact centers, it enables call recordings to be converted into transcripts that can be read, searched, and analyzed on a large scale.
What is the difference between batch transcription and live transcription?
In batch mode (asynchronous), recordings are transcribed after the call, in batches: this is the mode used for background analysis and large volumes. In live mode (synchronous), the text is generated during the conversation to detect signals and trigger alerts without waiting for the call to end.
In which languages can calls be transcribed?
Cross-Mining transcribes and analyzes conversations in more than 30 languages. Multilingual conversations can then be translated to standardize the analysis.
Can the transcription be hosted on-premises?
Yes. Cross-Mining is available as SaaS or on-premises, and the transcription engine is selected from among the analytics connectors based on your security and compliance requirements. See alsothe anonymization of personal data.
What do we do with the transcripts once they've been produced?
They feed into Cross-Mining 'sconversation analysis (sentiment, categorization, summaries, metrics), evaluations of Quality Monitoring and CRM Dataviz dashboards.
Other Speech Analytics features
Want to work with us?
Contact us
