Skip to content
/Product Updates
Rasa x Audiocodes Partnership

January 17th, 2023

Let your brand speak for itself: Introducing Voice at Rasa with Rasa Pro 3.4

  • portrait of Lauren Goerz

    Lauren Goerz

The legacy of the traditional IVR solution is ubiquitous. Many of us grew up navigating menu-based (DTMF) IVR Interfaces. Drawing from this shared experience, we all know the drill - Dial the number, “Press two for a one-size-fits-all solution,” and wait in a long phone queue with accompaniment by a jazz quartet.

Despite its bad rep, voice-based customer service is here to stay and remains one of the largest channels for customer support. The prevalence of these systems, and the universally poor experience and long wait times mean that the industry is ripe for change. It presents an opportunity for your business to leverage natural language technology to understand your users’ needs better and provide faster, contextual customer service over the phone that is genuinely superior to incumbent IVR solutions.

Why partner with Rasa for your voice solution

Compared to text-based customer support, voice-based customer support has no visual design affordances. This means that you don’t have the ability to guide your end-users with buttons, colorful carousels, or images as you do in text-based chat and that every triage question you ask costs more words per minute, lengthening the average handling time. With this in mind, it is imperative to partner with a vendor like Rasa that has strong contextual natural language understanding capabilities. Otherwise, you end up enabling a bot to deliver the same triage questions formerly delivered by IVR menus and there is not much won.

With Rasa, you can leverage our flexible NLP pipeline and customizable machine-learning capabilities to ensure your voice assistants are more successful right off the bat. When you partner with us to bring your brand voice to life, we can help you solve user problems faster and better than you can with less flexible voice solutions.

Introducing Rasa Pro 3.4: Our next step in the voice direction

With the most recent release of Rasa Pro 3.4, we have partnered with AudioCodes (A leading vendor of Voice over IP and VoIP Data Networking Products) to introduce a new voice offering for Rasa.

AudioCodes has developed a voice gateway (Voice AI Connect), a middleware that enables the deployment of Rasa AI Assistants over the phone, as well as some other key capabilities:

  • Connectivity with third-party STT (Speech-to-Text) /TTS (Text-to-Speech) vendors, as well as Audiocodes’ own STT and TTS capabilities for natural language processing
  • Interconnectivity with any contact center or SIP trunk
  • High-Quality voice and reduced voice latency
  • Control telephony actions from Rasa Pro such as escalation to human agents, play prompts, recording, Barge-in, and outbound calls
  • Ability to pass on contextual conversation metadata in the case of an escalation to a human agent

Additional Enhancements in Rasa Pro 3.4:

  • Support for Python 3.10: Support for Python 3.10 reduces vulnerabilities, and includes better error messaging, structural pattern matching, parenthesized context managers, and static typing improvements.
  • Operating System Update: The base OS version in Rasa OSS/Rasa Plus is now upgraded to Ubuntu 22.04
  • Authentication Option: Now you have the option for Symmetrical (Authenticated on passphrase) or Asymmetrical (authenticated based on both private and public key) encryption.
  • Websocket metadata support: Pass metadata onto the Rasa subsystem from websocket in order to create a more personalized experience.
  • Control over your log formatting: New custom log configuration gives you the freedom to have more control over the format of your logs

Next Steps

Are you interested in taking the next step of your voice journey with Rasa? Reach out to connect with an expert. In the meantime check out our AudioCodes channel connector documentation.