Hello Kaian (and the community),
I’m genuinely excited about the opportunity IvozProvider brings to the open-source VoIP ecosystem, and I do my best to promote it wherever possible.
One topic that has come up for me recently is the use of AI within IvozProvider. Are there any current or future plans to introduce AI-related capabilities into the platform?
I’ve been experimenting with a couple of Asterisk-based open-source Voice AI projects lately, namely AgentVoiceResponse (agentvoiceresponse.com) and Asterisk AI Voice Agent (https://github.com/hkjarral/Asterisk-AI-Voice-Agent). Both rely heavily on Asterisk AMI and ARI. I haven’t yet gone deep into the Asterisk internals used by IvozProvider, but from my understanding it primarily uses PHP-based AGI.
I’m curious about the best way to start experimenting with these AI projects inside an IvozProvider test environment running in my homelab, or whether there are recommended integration points for this kind of work.
Additionally, are there any plans on the roadmap for:
Thank you in advance for your time and guidance.
br,
airsay
Hi Airsay,
All our AI-related testing in ivozprovider has been carried out using external software connected to the platform through Friends/RetailAccounts, such as LiveKit or LiveKit-based platforms (ElevenLabs, RetellAI, etc.). At this time, we are not planning to integrate AI bots or agents directly into ivozprovider. Instead, our approach is to provide integration points for third-party platforms that already have expertise in these areas.
Regarding the other features, none of our clients have shown interest in SMS support, and none of them are based in the US. As a result, STIR/SHAKEN has not been requested or discussed.
Best Regards,
Kaian