Conversational IVR
To do this, Expertflow CX has a Session Border Controller (to re-name) to embed any speech engine for speech recognition, NLU for recognizing customer's intent and for scripting. This determines what to say, and how to speak back to the customers. The information received from Conversation IVR will be stored as Conversation Data in slots in future.
Business Features
| Features | Benefits |
|---|---|
| Better Enquiry Resolution | Using AI-powered Natural Language Understanding and machine learning, CX Conversational IVR system directs customer calls to the best available agent, or even completely resolve the customer query without involving any agent. |
| Agent Assistant Mode | Has the capability to work in Agent Assistant Mode by routing the queries to the right agent. |
| Integration with any type of AI Engine | Is capable to work with any type of AI engine such as RASA, Amazon Lex for voice biometrics, sentiment analysis. |
| Integration with Speech Engines | Provides automated speech recognition by integrating with speech engines such as Microsoft Cognitive Services, Amazon Lex, Google DialogFlow with tested languages of English and Urdu. Any other language support can be available on request. |
| Initial Call Controls | Provides initial call controls such as call accept, call route, call end. |
| Scalability | Can handle multiple simultaneous interactions by its automated system and prevent surges affecting the quality. |
| Round the Clock Availability | Is available round the clock, 24/7, 365 days a year without making customers worry about traditional business hours. |
| Data Availability for Analytics | Ensures large amount of data available for analysis by leveraging NLU capabilities. |
Integration with Speech Engines
Expertflow CX provides integration with the following speech engines:
- Amazon Lex
- Microsoft Cognitive Services and
- Google DialogFlow
which provides additional capabilities (along with voice, and chatbots) such as voice and face biometrics, emotion detection, transcription, and translation/ transcription during or after a call.
The speech services can be available in any language but the following languages have been tested as of now:
- English
- Urdu
Speech services in any language can be made available on demand.
Integration with NLU
Expertflow CX provides integrations with:
- RASA
- Amazon Lex
Architecture
Business/Communication Flow (depends upon the architecture)
Build Your Own Conversational IVR
To set up a simple conversational IVR, make an application in NodeRED or Node.js.
NodeRED
Node-RED is a flow-based development tool for visual programming developed originally by IBM for wiring together hardware devices, APIs and online services as part of the Internet of Things. Node-RED provides a web browser-based flow editor, which can be used to create JavaScript functions.
Jambonz provide us NodeRED plugin which we easily install and made flows.
For Installing Jambonz plugin visit https://www.jambonz.org/docs/client-sdks/node-red/
Node.js SDK
The Jambonz Node.js SDK is one of the simplest and most powerful way to build Jambonz Applications. Here, more controls are observed than NodeRED. More customization and flows are available.
For creating an application using Node.js SDK, visit https://www.jambonz.org/docs/client-sdks/nodejs-sdk/
Pre-requisites
Single Instance Hardware Requirements
For every single instance, 3 Debian servers are required to deploy. One for the feature server and 2nd for RTP server.
Feature Server | RTP Server | Node Server | |
|---|---|---|---|
Hardware Sizing | |||
For 300 to 400 simultaneous calls at the feature server and 600 to 800 simultaneous calls at RTP (Real-time Transport Protocol) server. | 4 vCPU, 8 GB RAM, 150 GB HDD | 4 vCPU, 8 GB RAM, 150 GB HDD | 2 vCPU, 4 GB RAM, 80 GB HDD |