pegachatbot.py

import google.generativeai as genai
import streamlit as st
from PIL import Image

# Set up Streamlit page configuration
st.set_page_config(page_title="Pega Tutor", page_icon="🎓", layout="wide")

# Load and display logo/image
image = Image.open("pega.jpeg")
col1, col2 = st.columns([1, 3])
with col1:
    st.image(image, width=250)  # Increased the width to 150
with col2:
    st.title("Pega Tutor Application")
    st.write("An expert AI-powered tutor to help with your Pega-related questions.")

# Configure API key
genai.configure(api_key="AIzaSyB-7cKMdUpA5kTccpNxd72IT5CjeSgSmkc")

# System prompt for the generative model
sys_prompt = """
You are an experienced Tutor with 20 years of professional expertise in the Pega Customer decision hub and pega systems expert.
Your role is to help students by answering their questions related to Pega in a very clear, simple, 
and easy-to-understand manner. Provide detailed explanations and use relatable examples to help 
illustrate your points effectively. If a student asks a question outside the scope of Pega, politely 
decline and remind them to ask questions only related to Pega platform.
"""

# Initialize the generative model
model = genai.GenerativeModel(model_name="models/gemini-1.5-flash", system_instruction=sys_prompt)

# Initialize session state to store chat history
if "chat_history" not in st.session_state:
    st.session_state.chat_history = []

# Function to generate response
def generate_response():
    user_prompt = st.session_state.user_prompt
    if user_prompt:
        with st.spinner("Generating answer..."):
            response = model.generate_content(user_prompt)
            st.session_state.response_text = response.text
            # Store user question and AI response in the chat history
            question_snippet = user_prompt[:30]  # Get the first 30 characters of the question
            st.session_state.chat_history.append((f"Q: {question_snippet}", user_prompt))
            st.session_state.chat_history.append((f"A: {question_snippet}", st.session_state.response_text))
    else:
        st.session_state.response_text = "Please enter a query before pressing Enter."


# Input from the user using chat_input
human_prompt = st.chat_input(" Message Pega ...")

if human_prompt:
    st.session_state.user_prompt = human_prompt
    generate_response()

# Display chat history with collapsible question/answer sections
st.sidebar.title("Chat History")
for idx, (label, message) in enumerate(st.session_state.chat_history):
    if "Q:" in label:
        with st.sidebar.expander(f"{label}"):
            st.markdown(f"**{label}**: {message}")
            # Find and display the corresponding answer
            answer_label = f"A: {label[2:]}"  # Get corresponding answer snippet
            answer_message = st.session_state.chat_history[idx + 1][1] if idx + 1 < len(st.session_state.chat_history) else ""
            st.markdown(f"**{answer_label}**: {answer_message}")

# Display the response
if 'response_text' in st.session_state:
    st.markdown("#### Tutor's Response:")
    st.write(f"🧑‍🏫: {st.session_state.response_text}")

# Display footer or additional help text
st.write("---")