- Joined
- Mar 22, 2024
- Messages
- 1
- Reaction score
- 0
Hello. This is my first time posting here, hopefully someone has an answer to my questions.
Our company has created an AI model that can answer complex questions with regards insurance (we are using OpenAI as our LLM but we provided our own knowledgebases). We already deployed this same model in our own e-commerce (in a form of chatbot) and whatsapp official number (so if a customer replies to our official whatsapp number, the AI bot will takeover the conversation).
Now, we are planning to deploy the same model to voice but I can't seem to find answers online.
Use Case:
1. Customer calls our hotline
2. AI Bot answers the call
3. Customer asks question (the way I imagine this in my head is to convert speech to text)
4. AI Bot understands the question (same text should be run in our AI model)
5. AI Bot gives response (AI model gives response in text format. Convert text to speech then stream in call)
Also, our AI Engineer suggested for me search about audio-stream support for 3CX (to get inputs from customer and give outputs to customer) but I can't seem to find any answers online.
If anyone is currently doing a project of the same, please do reach out to me.
Our company has created an AI model that can answer complex questions with regards insurance (we are using OpenAI as our LLM but we provided our own knowledgebases). We already deployed this same model in our own e-commerce (in a form of chatbot) and whatsapp official number (so if a customer replies to our official whatsapp number, the AI bot will takeover the conversation).
Now, we are planning to deploy the same model to voice but I can't seem to find answers online.
Use Case:
1. Customer calls our hotline
2. AI Bot answers the call
3. Customer asks question (the way I imagine this in my head is to convert speech to text)
4. AI Bot understands the question (same text should be run in our AI model)
5. AI Bot gives response (AI model gives response in text format. Convert text to speech then stream in call)
Also, our AI Engineer suggested for me search about audio-stream support for 3CX (to get inputs from customer and give outputs to customer) but I can't seem to find any answers online.
If anyone is currently doing a project of the same, please do reach out to me.