Skip to content

Latest commit

 

History

History
197 lines (171 loc) · 6.61 KB

README.md

File metadata and controls

197 lines (171 loc) · 6.61 KB
Raw

VitalPBX - AI Agent with OpenAI ChatGPT, Whisper and Microsoft Azure AI Speech (TTS)

VitalPBX AGENT AI

Necessary Resources

  1. OpenAI Account (https://platform.openai.com/apps).
  2. Microsoft Azure Account (https://azure.microsoft.com/en-us/products/ai-services/text-to-speech)
  3. VitalPBX 4

Installing dependencies

  apt update
  apt install python3 python3-pip
  pip install azure-cognitiveservices-speech
  pip install pyst2
  pip install pydub
  pip install python-dotenv==0.21.0
  pip install langchain==0.0.331rc2
  pip install pypdf==3.8.1
  pip install docx2txt==0.8
  pip install openai
  pip install chromadb==0.3.22
  pip install tiktoken==0.4.0
  pip install markdown

Install from script

Download the script

  wget https://raw.githubusercontent.com/VitalPBX/vitalpbx_agent_ai_chatgpt/main/install.sh

Give execution permissions

  chmod +x install.sh

Run the script

  ./install.sh

Create .env file

Goto AGI directory

  cd /var/lib/asterisk/agi-bin/

Creating .env

  nano .env

Copy the following content and add the APIS Key.

OPENAI_API_KEY = "sk-"
AZURE_SPEECH_KEY = ""
AZURE_SERVICE_REGION = "eastus"
PATH_TO_DOCUMENTS = "/var/lib/asterisk/agi-bin/docs/"
PATH_TO_DATABASE = "/var/lib/asterisk/agi-bin/data/"

Create voice guides

Goto AGI directory

  cd /var/lib/asterisk/agi-bin/

The format to record a prompt is as follows: ./record-prompt.py file-name "Text to record" language
file-name --> file name if extension mp3, remember that in the Agent AI script, the welcome audio is: welcome-en (English), welcome-es (Spanish), and the wait audio is: wait-en (English), and wait-es (Spanish).
languaje --> could be "en-US" or "es-ES"
If you want to add more languages, you must modify the scripts

Below we show an example of how you should use the script to record the prompt.

./record-prompt.py wait-en "Just a moment, please. We're fetching the information for you." "en-US"
./record-prompt.py welcome-en "Welcome to the Artificial intelligence Assistant. Please ask your question after the tone." "en-US"
./record-prompt.py short-message-en "Your question is too short. Please provide more details." "en-US"
./record-prompt.py anything-else-en "Can I assist you with anything else?" "en-US"  
./record-prompt.py wait-es "Un momento, por favor. Estamos buscando la información para ti." "es-ES"
./record-prompt.py welcome-es "Bienvenido al Asistente de Inteligencia Artificial, Haga su pregunta después del tono." "es-ES"
./record-prompt.py short-message-es "Tu pregunta es demasiado corta. Por favor, proporciona más detalles." "es-ES"
./record-prompt.py anything-else-es "¿Hay algo más en lo que pueda ayudarte?" "es-ES"

Testing Embedding

To test the functionality of our AI Agent with the Embedding option, run the following script:
1.- Upload the document to the /var/lib/asterisk/agi-bin/docs folder with the information to use for the query with ChatGPT-Embedded
2.- To transfer this document to a Vector database (ChromaDB), proceed to execute the following command.

  cd /var/lib/asterisk/agi-bin/
  ./vpbx-embedded-docs.py

3.- Now execute the following command and we proceed to ask any question in reference to the document that we previously uploaded.

  cd /var/lib/asterisk/agi-bin
  ./chatbot.py

Testing call from VitalPBX

To ask ChatGPT questions: Dial *778 for English or *779 for Spanish
For queries obtained from custom documentation, first upload the PDF document in the docs folder, then run from the console the command:

  cd /var/lib/asterisk/agi-bin/ 
  ./vpbx-embedded-docs.py

To ask ChatGPT-Embedded questions: Dial *888 for English or *889 for Spanish

View in real time

It is possible to see the question and answer in real time on a web page, for which we are going to follow the following procedure.

Install prerequisites

pip install websocket-client
pip install asyncio

We are going to copy the chatserver.py file to the folder we want (It could be /var/lib/asterisk/agi-bin/).

  cd /var/lib/asterisk/agi-bin/
  wget https://raw.githubusercontent.com/VitalPBX/vitalpbx_agent_ai_chatgpt/main/chatserver.py
  chamod +x chatserver.py

Edit the file you just downloaded (chatserver.py)

  cd /var/lib/asterisk/agi-bin/
  nano chatserver.py

Replace the following lines with your IP or Domain if you use SSL.
If you use ssl with a valid domain remember to change ws to wss.

start_server = websockets.serve(echo, 'Your_IP_or_Domain', 3001)
print("WebSocket server started on ws://Your_IP_or_Domain:3001")

Now we will proceed to create the service

  cd /etc/systemd/system/
  nano vpbxagentai.service

Copy and paste the following content

[Unit]
Description=Agent AI
After=network.target

[Service]
ExecStart=/usr/bin/python3 /var/lib/asterisk/agi-bin/chatserver.py
Restart=always
User=root
Group=root
Environment=VariableDeEntorno=valor
WorkingDirectory=/var/lib/asterisk/agi-bin

[Install]
WantedBy=multi-user.target

Enable and start the service

systemctl enable vpbxagentai
systemctl start vpbxagentai
systemctl status vpbxagentai

Now we are going to download the chat.html file and copy it to the /usr/share/vitalpbx/www folder

  cd /usr/share/vitalpbx/www
  mkdir chatview
  cd chatview
  wget https://raw.githubusercontent.com/VitalPBX/vitalpbx_agent_ai_chatgpt/main/chatview/index.html
  wget https://raw.githubusercontent.com/VitalPBX/vitalpbx_agent_ai_chatgpt/main/chatview/vpbx-agent-ai-m.png

For HTTPS: wss, for HTTP: ws. If you are not going to use SSL just enter the IP of your server, otherwise leave ${location.hostname}.

Finally, edit the files vpbx-agent-ai-embedded.py and vpbx-agent-ai.py and uncomment everything related to sending messages via websocket.

To see the chat in real time, run the url of your VitalPBX:
For example:
http://mypbxurl/chatview
or
https://mypbxurl/chatview

Note

Remember to unblock port 3001 or the one you decided to use in the VitalPBx firewall as in any other firewall that VitalPBX has in front of you.
To make sure everything is fine, we can run the following command.

  netstat -tuln | grep 3001

And it would have to return the following to us:

tcp        0      0 192.168.57.50:3001       0.0.0.0:*               LISTEN     
tcp        0      0 127.0.1.1:3001           0.0.0.0:*               LISTEN

192.168.57.50 is our public or private IP.