Initial Llamaindex commit

[AIMikav]

This PR introduces the initial implementation of a multi-agent framework using LlamaIndex with the following updates:

1. FastAPI Backend: RESTful API for insurance score generation and approval using LLM.
2. Streamlit Frontend: Simple UI for user interaction with the agent.
3. Environment Variable Support: Configurable LLM model via LLAMA_MODEL_NAME.
4. README Update: Instructions for setting up FastAPI, Streamlit, environment variables, and added image documentation.

[MichaelClifford]

Its a good practice to not actually include the .env file in your repo as it can leak credentials. I recommend adding a .env.example with dummy values for the parameters, and adding a .gitignore that includes .env so it doesn't get tracked by git.

[tumido]

Just a code quality/code smell nit. 🙂 It is good to keep your code clean so it's easier to maintain. It reduces possibility of bugs, allows others to orient themselves in the codebase much faster etc and in the end leads to fewer errors...

[tumido]

Just a nit for code quality. The Zen of Python says: Flat is better than nested.

Suggested change

	if st.button("Submit"):
	if user_input:
	# Call FastAPI backend with the user input
	url = "http://127.0.0.1:8000/process_input/"
	response = requests.post(url, json={"text": user_input})
	if response.status_code == 200:
	result = response.json()
	st.write(f"Generated Insurance Score: {result['response']}")
	st.write(f"Approval Status: {result['approval_status']}")
	else:
	st.write("There was an error processing the request.")
	else:
	st.write("Please enter some information about your driving history.")
	if not st.button("Submit"):
	return
	if not user_input:
	st.write("Please enter some information about your driving history.")
	return
	...
	if response.status_code != 200:
	st.write("There was an error processing the request.")
	return
	...

[AIMikav]

Deleted .env file and included .gitignore and .env.example, Also changed the streamlit_ui.py script with less nesting.