import execution.euron_streamlit as st
from src.helper import download_hugging_face_embeddings
from langchain_pinecone import PineconeVectorStore
from langchain.chains import create_retrieval_chain
from langchain.chains.combine_documents import create_stuff_documents_chain
from langchain_core.prompts import ChatPromptTemplate
from src.prompt import *
from src.euron_chat import EuronChatModel
from dotenv import load_dotenv
import os
# Load environment variables
load_dotenv()
pinecone_api_key = os.getenv("PINECONE_API_KEY")
euron_api_key = os.getenv("EURON_API_KEY")
if not pinecone_api_key or not euron_api_key:
st.error("Missing keys in environment")
st.stop()
# ✅ Minimal CSS – Only apply Fira Code font globally + Sidebar styling
st.markdown("""
""", unsafe_allow_html=True)
# Sidebar Menu
with st.sidebar:
# st.title("🌟 Navigation")
st.image("static/icon.png", width=120) # Image in sidebar
st.markdown("---")
menu = st.radio("Go to", ["About Me", "Chatbot"], label_visibility="hidden")
st.markdown("---")
st.caption("© 2025 Shanin Hossain")
# st.markdown("""
#
# """, unsafe_allow_html=True)
# Initialize session state
if "rag_chain" not in st.session_state:
st.session_state.rag_chain = None
st.session_state.embeddings = None
st.session_state.retriever = None
if "messages" not in st.session_state:
st.session_state.messages = [] # Chat history
def initialize_rag():
try:
if st.session_state.rag_chain is None:
st.session_state.embeddings = download_hugging_face_embeddings()
index_name = "portfolio"
docsearch = PineconeVectorStore.from_existing_index(
index_name=index_name,
embedding=st.session_state.embeddings
)
st.session_state.retriever = docsearch.as_retriever(
search_type="similarity", search_kwargs={"k": 3}
)
chatModel = EuronChatModel()
prompt = ChatPromptTemplate.from_messages(
[
("system", system_prompt),
("human", "{input}"),
]
)
question_answer_chain = create_stuff_documents_chain(chatModel, prompt)
st.session_state.rag_chain = create_retrieval_chain(
st.session_state.retriever, question_answer_chain
)
except Exception as e:
st.error(f"Error initializing RAG: {str(e)}")
st.stop()
# About Me Section
# About Me Section
if menu == "About Me":
st.write("Hey there 👋")
st.title("👨 I'm Shanin Hossain")
st.markdown("""
I'm an AI Engineer & Research Assistant passionate about:
- Machine Learning, Deep Learning, and Generative AI
- Computer Vision & Natural Language Processing
- Healthcare Informatics and Medical Imaging
📌 I have worked on multiple AI projects, including OCR, Retrieval-Augmented Generation (RAG), and hybrid graph networks.
""")
st.success("👉 Navigate to **Chatbot** in the sidebar to chat with me!")
st.header("My Projects", divider="gray")
# --- Project Data ---
projects = [
{
"name": "Brain Glioma Grading System",
"description": "Developed a hybrid graph neural network to grade glioma tumors from medical imaging data.",
"tech_stack": ["PyTorch Geometric", "Graph Neural Networks", "Medical Imaging", "Python"]
},
{
"name": "OCR Automation",
"description": "Built an OCR pipeline for document image understanding and text extraction.",
"tech_stack": ["YOLOv8", "OpenCV", "Tesseract OCR", "FastAPI"]
},
{
"name": "Portfolio Chatbot",
"description": "Created a RAG-powered chatbot integrated with Pinecone & custom embeddings for Q&A over portfolio data.",
"tech_stack": ["Streamlit", "LangChain", "Pinecone", "Hugging Face"]
},
]
# --- Render Projects with Badges ---
for project in projects:
with st.container():
st.subheader(project["name"])
st.write(project["description"])
# Create badges for each tech in stack
badges_html = " ".join([
f"{tech}" for tech in project["tech_stack"]
])
st.markdown(
f"{badges_html}
",
unsafe_allow_html=True
)
st.markdown("---")
# --- Badge Styling ---
st.markdown("""
""", unsafe_allow_html=True)
st.header("Publications", divider="gray")
# --- Publication Data ---
publications = [
{
"title": "Automated Detection of Age-Related Macular Degeneration (AMD) Using Deep Learning",
"venue": "Journal of Medical Imaging & Health Informatics, 2023",
"description": "Published a deep learning-based pipeline to detect age-related macular degeneration from retinal images.",
"link": "https://doi.org/xxxxxx"
},
{
"title": "Using Hyperdimensional Computing to Extract Features for the Detection of Type 2 Diabetes",
"venue": "Conference on Health Informatics, 2024 (Under Review)",
"description": "Explored hyperdimensional computing techniques to improve detection of Type 2 Diabetes from clinical data.",
"link": ""
}
]
# --- Render Publications ---
for pub in publications:
with st.container():
st.subheader(pub["title"])
st.caption(pub["venue"])
st.write(pub["description"])
if pub["link"]:
st.markdown(f"[🔗 View Publication]({pub['link']})")
st.markdown("---")
# Chatbot Section
elif menu == "Chatbot":
st.title("🤖 Shanin Chatbot")
st.write("Ask me anything about my portfolio!")
# Display previous messages
for msg in st.session_state.messages:
with st.chat_message(msg["role"]):
st.write(msg["content"])
# Input box
if prompt := st.chat_input("Type your question..."):
# Add user message
st.session_state.messages.append({"role": "user", "content": prompt})
with st.chat_message("user"):
st.write(prompt)
# Initialize RAG if needed
if st.session_state.rag_chain is None:
with st.spinner("Initializing RAG pipeline..."):
initialize_rag()
try:
with st.spinner("Generating response..."):
response = st.session_state.rag_chain.invoke({"input": prompt})
answer = response["answer"]
# Add assistant response
st.session_state.messages.append({"role": "assistant", "content": answer})
with st.chat_message("assistant"):
st.write(answer)
except Exception as e:
st.error(f"Error processing request: {str(e)}")
#
#
# 
#
#
# 
#
#