Advanced AI for Technology Sales Brazilian Portuguese Cold Call AI
Budget: $250 – $750 USD
1. PROJECT OBJECTIVE: (similar to www.air.ai)
To develop an artificial intelligence capable of making phone calls.
- we need the AI to call the lead and offer the product, being able to converse
and process information with the lead for approximately 1 to 40 minutes if
necessary.
- we need the conversation to be fluid and interactive, giving the impression of a
person speaking.
- provide 3 different female voices and 3 different male voices.
- during the call, the AI will schedule a meeting, being able to access a schedule
per product and schedule meetings with the lead.
- note: We will have a series of products, and each product will have 1 to
10 fixed schedules.
- the AI needs to access the schedule of the product it is prospecting and
schedule on the agenda of this product.
- after each conversation, the AI needs to learn more about the types of
customers and the questions that may arise and improve itself based on positive
results.
- generate a conversation history between the AI and the lead:
- text-based history
- audio-based history (if possible)
a) How we intend to train the AI:
i) by accessing our Bubble account, the AI will have access to the products
offered. We intend for it to read, listen, watch, and learn about the
products we offer and make calls offering these products.
ii) throughout each positive call, the AI will schedule a meeting with the
lead.
1) if the lead does not want to schedule calls, the AI will classify the
lead as "not interested."
2) if the lead does not answer the call, the AI will classify it as
"unavailable."
b) What we will receive from the development:
i) AI described above
A.I. FOR COLD CALL
ii) AI code (API) via GitHub
iii) Connection of the API with the Bubble.io system we use.
iv) Setup of a field for phone numbers for the AI to make calls through these
configured numbers. (N to N)
v) Meeting to explain how the API will work on Bubble, and how we will
configure the AI to learn about the products, make calls, and schedule
meetings
observations: the front end is already ready in bubble end I have a project on github that you can use to do this work.
I only pay within the platform
Tool & Technologies Used
Python3
FastAPI
SQLModel ORM
OpenAI (for answer generation)
AWS Transcribe (for transcription)
Twilio (for calling)
Elevenslab (for text to voice conversion)
Core Features
JWT Authentication system in FastAPI
Handle multiple connections of the system users in websockets
System user can choose a voice before making a call to the customer
User can see and listen all the available voices and choose one of them
User can upload an excel file of customer contacts for the campaign call (one call at a time)
To develop an artificial intelligence capable of making phone calls.
- we need the AI to call the lead and offer the product, being able to converse
and process information with the lead for approximately 1 to 40 minutes if
necessary.
- we need the conversation to be fluid and interactive, giving the impression of a
person speaking.
- provide 3 different female voices and 3 different male voices.
- during the call, the AI will schedule a meeting, being able to access a schedule
per product and schedule meetings with the lead.
- note: We will have a series of products, and each product will have 1 to
10 fixed schedules.
- the AI needs to access the schedule of the product it is prospecting and
schedule on the agenda of this product.
- after each conversation, the AI needs to learn more about the types of
customers and the questions that may arise and improve itself based on positive
results.
- generate a conversation history between the AI and the lead:
- text-based history
- audio-based history (if possible)
a) How we intend to train the AI:
i) by accessing our Bubble account, the AI will have access to the products
offered. We intend for it to read, listen, watch, and learn about the
products we offer and make calls offering these products.
ii) throughout each positive call, the AI will schedule a meeting with the
lead.
1) if the lead does not want to schedule calls, the AI will classify the
lead as "not interested."
2) if the lead does not answer the call, the AI will classify it as
"unavailable."
b) What we will receive from the development:
i) AI described above
A.I. FOR COLD CALL
ii) AI code (API) via GitHub
iii) Connection of the API with the Bubble.io system we use.
iv) Setup of a field for phone numbers for the AI to make calls through these
configured numbers. (N to N)
v) Meeting to explain how the API will work on Bubble, and how we will
configure the AI to learn about the products, make calls, and schedule
meetings
observations: the front end is already ready in bubble end I have a project on github that you can use to do this work.
I only pay within the platform
Tool & Technologies Used
Python3
FastAPI
SQLModel ORM
OpenAI (for answer generation)
AWS Transcribe (for transcription)
Twilio (for calling)
Elevenslab (for text to voice conversion)
Core Features
JWT Authentication system in FastAPI
Handle multiple connections of the system users in websockets
System user can choose a voice before making a call to the customer
User can see and listen all the available voices and choose one of them
User can upload an excel file of customer contacts for the campaign call (one call at a time)