PolyAI puts its Dialog-RSN-1 voice model on Amazon SageMaker
The model hears caller audio directly and works in the AWS account of the customer.
PolyAI makes voice agents for phone calls. It put its model Dialog-RSN-1 on AWS Marketplace and Amazon SageMaker JumpStart. JumpStart is a catalog of models that a team can start in a small number of clicks. The model gets the audio of a live phone call and gives an answer in one step. PolyAI gave no test results.
How it works
Many voice systems change the voice to text first, and a different model uses the text. Dialog-RSN-1 does not do this. It reads the audio and finds when to speak. The customer keeps the control of phone routing and its business rules.
The numbers
There are no test results. PolyAI tells that the model is in use on live calls. These limits apply:
- Audio in must be 16 or 24 kHz.
- A connection stays open for a maximum of 30 minutes.
- The model uses English only.
What is not known
The source gives no price and no speed numbers. PolyAI tells that the model is fast and accurate, but it shows no proof. It also tells that the data stays in the account of the customer.
Insight: AWS is becoming the one door for voice AI
On 11 November 2026, Amazon added billing by use for live voice sessions, and Deepgram is the first partner. Then PolyAI put a live voice model on the same shop, billed through the AWS account of the customer. This reads as a move by AWS to be the one door for voice AI vendors. It is also the one bill for buyers. Next, look for more voice vendors that use this billing.
Sources
- PolyAI Dialog-RSN-1 is now on Amazon SageMaker JumpStartaws.amazon.com
Posted