This is the PoC application for showcasing a rapid development of a simple multi-agent domain specific chatbot.
Uses Groq and Streamlit and llama3 models to generate response.
Uses sqlite3 for database stuff.
Conveyancing in the UK
intent_checker_agent
checks for intent of the user query - internal interactionanswer_responder_agent
responds to the end user - user exposuredb_agent
queries database and parses into human readable format.
GROQ_API_KEY=
USER_ID=stargazer94
Python==3.13.0
pip==24.2
- Clone this repository
- Install python and upgrade pip if necessary
- Install libraries mentioned in the
requirements.txt
- Create an
.env
file and populate your Groq api key underGROQ_API_KEY
. - Run
db_init.py
that should create some sample data for conveyancing demo. Make sure this createschatbot.db
file in the existing repository. - Open the command in the current folder and run the command
streamlit run ./chatbot.py
to start the server.