Hey fellow WEMers,

How's your week going?

Did you know there are lot of features that use AI in Speech and Text Analytics!

These include:

• Voice Transcription
• Sentiment Analysis
• Agent Empathy Analysis
• Topic Miner
• Sensitive Data Masking 

And not only that - check out what is also on the Roadmap 🙌🙌🙌🙌

• Automated Interaction Insights
• Copilot Translate
• Dictionary Management

·       To put a bit more context into these and what they are, let me try to explain in my own way😉

What is Voice Transcription?

With Voice Transcription - get complete and accurate transcriptions of every interaction, turn all customer engagements into data that you can use to improve their experiences and your business outcomes.

All Deployment Types Supported (GCV, BYOCC, BYOCP)

Genesys Native Transcription Supported Languages:

•       English (US, UK, AU, ZA, IN), Spanish (ES, LATAM),

•       French (FR, CA), Dutch, Italian, German, Polish

•       Portuguese (PT, BR), Korean, Japanese, Arabic

Extended Voice Transcription (using Azure STT) Languages (40+ Dialects)

Near Real Time Transcripts through Notification API

What is Sentiment Analysis?

It examines interaction language and gains a deeper customer understanding. Get a better grasp of customer attitudes and perceptions, providing vital data about your product, service agents and areas needing enhancement.

Adapt Sentiment Results by Providing Feedback

• You can add a phrase where you feel that the system, should have marked that phrase as Positive, Negative or Neutral
• Subsequently, when that phrase is encountered, Sentiment Analysis will use the Feedback provided to override anything found by the built-in Sentiment Analysis model.

What is Agent Empathy?

It measures agent empathy / unhelpful behavior towards customers.

With it you can detect where in the interaction agents are expressing empathy and unhelpful behavior to customers.

You can then use this information for analysis and coaching opportunities.

And search for interactions with unhelpful behavior to more easily find opportunities to train and improve agents.

What is Topic Miner?

With the use of AI - discover new Topics or Phrases to add.

Use machine learning (i.e., key-phrase extraction), to mine transcripts (voice, chat, or messages), to find common topics of conversation along with related phrases to help you create company specific topics.

The selection of Data Source for mining can be:

• Date Range
• Media Type
• Participant(s)
• Queues

And the results show:

# and % of Conversations where Topic is found

# of unique sentences and phrases discovered

This allows the creation of a new Topic or merging with an Existing Topic.

What is Sensitive Data Masking?

It Uses AI to Find PCI and PII Data:

Uses Named Entity Recognition (NER) to try to find on a best effort basis:

•      PCI Data: Credit Card Numbers, Expiration Dates, CVV / CID

•      PII Data : Person Name, Postal Codes, Phone Numbers, Locations, etc.

It Masks / Redacts Sensitive Data:

Indicators to show if Sensitive Data Masking was run and outlines where Sensitive Data was found in Transcripts.  Audio and Transcripts are masked on the fly when accessed.  Original remains unmodified.

It has Permissions Control Access

Users with elevated access permissions can still access the original, unredacted recording and transcript.

Bulk and Batch Export Jobs block access to recordings where Sensitive Data was detected if elevated permissions not granted.

Note that this feature does NOT contribute to your solution being PCI Compliant.  This only aims for better data stewardship.

Well, I do hope that gives you an insight of what we have now - must admit I didn't realise AI was that involved!!

Oh yes there is, here's a little more about whats on the roadmap…….

Automated Interaction Insights where you will be able to Utilize llms to Auto Generate Key Insights

and

Copilot Translate that will utilize advancements in Machine Translation to aid Multi-Lingual Organizations

and

Dictionary Management for enhancing transcription accuracy

I don't have much more info than this but keep an eye out for more information as we learn more here on the WEM Community!

#SpeechandTextAnalytics

------------------------------
Tracy
Genesys
------------------------------