Upload 16 files

.gitattributes

@@ -0,0 +1,35 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text

.streamlit/config.toml

@@ -0,0 +1,4 @@
+[server]
+enableCORS = false
+enableXsrfProtection = false
+maxUploadSize = 200

Dockerfile

@@ -0,0 +1,21 @@
+FROM python:3.11-slim
+
+WORKDIR /app
+
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    curl \
+    git \
+    && rm -rf /var/lib/apt/lists/*
+
+COPY requirements.txt ./
+COPY src/ ./src/
+COPY . .
+
+RUN pip3 install -r requirements.txt
+
+EXPOSE 8501
+
+HEALTHCHECK CMD curl --fail http://localhost:8501/_stcore/health
+
+ENTRYPOINT ["streamlit", "run", "streamlit_app.py", "--server.port=8501", "--server.address=0.0.0.0"]

README.md

@@ -0,0 +1,20 @@
+---
+title: Demo 1
+emoji: 🚀
+colorFrom: red
+colorTo: red
+sdk: docker
+app_port: 8501
+tags:
+- streamlit
+pinned: false
+short_description: for learning
+license: mit
+---
+
+# Welcome to Streamlit!
+
+Edit `/src/streamlit_app.py` to customize this app to your heart's desire. :heart:
+
+If you have any questions, checkout our [documentation](https://docs.streamlit.io) and [community
+forums](https://discuss.streamlit.io).

requirements.txt

@@ -0,0 +1,13 @@
+langchain
+langchain-community
+sentence-transformers
+langchain-huggingface
+langchain-openai
+langgraph
+openai
+faiss-cpu
+pydantic
+python-dotenv
+requests
+streamlit
+pypdf

src/config/config.py

@@ -0,0 +1,20 @@
+# src/config.py
+import os
+
+# Embeddings
+EMBEDDING_MODEL = "sentence-transformers/all-MiniLM-L6-v2"
+EMBEDDING_DEVICE = "cpu"
+NORMALIZE_EMBEDDINGS = True
+
+# LLM
+LLM_MODEL = "openai/gpt-oss-120b:free"
+OPENROUTER_BASE_URL = "https://openrouter.ai/api/v1"
+OPENROUTER_API_KEY = os.getenv("OPENROUTER_API_KEY")
+
+# Text Splitter
+CHUNK_SIZE = 500
+CHUNK_OVERLAP = 100
+
+# Retriever
+MMR_LAMBDA = 0.25
+K_OFFSET = 2

src/core/embeddings.py

@@ -0,0 +1,17 @@
+from langchain_huggingface import HuggingFaceEmbeddings
+from src.config.config import (
+    EMBEDDING_MODEL,
+    EMBEDDING_DEVICE,
+    NORMALIZE_EMBEDDINGS
+)
+
+def load_embeddings():
+    try:
+        return HuggingFaceEmbeddings(
+            model_name=EMBEDDING_MODEL,
+            model_kwargs={"device": EMBEDDING_DEVICE},
+            encode_kwargs={"normalize_embeddings": NORMALIZE_EMBEDDINGS}
+        )
+    except Exception as e:
+        raise RuntimeError(f"Failed to load embeddings: {e}")
+

src/core/graph_state.py

@@ -0,0 +1,8 @@
+from typing import List, TypedDict
+from langchain_core.documents import Document
+
+class GraphState(TypedDict):
+    question: str
+    context: List[Document]
+    answer: str
+

src/core/llm.py

@@ -0,0 +1,17 @@
+# src/llm.py
+from langchain_openai import ChatOpenAI
+from src.config.config import (
+    LLM_MODEL,
+    OPENROUTER_BASE_URL,
+    OPENROUTER_API_KEY
+)
+
+def load_llm():
+    if not OPENROUTER_API_KEY:
+        raise EnvironmentError("OPENROUTER_API_KEY not set")
+
+    return ChatOpenAI(
+        model=LLM_MODEL,
+        base_url=OPENROUTER_BASE_URL,
+        api_key=OPENROUTER_API_KEY
+    )

src/exceptions.py

@@ -0,0 +1,8 @@
+class DocumentProcessingError(Exception):
+    pass
+
+class VectorStoreNotInitializedError(Exception):
+    pass
+
+class LLMInvocationError(Exception):
+    pass

src/rag_graph.py

@@ -0,0 +1,92 @@
+# src/rag_graph.py
+from langgraph.graph import StateGraph, END
+from langgraph.checkpoint.memory import MemorySaver
+from langchain_core.prompts import ChatPromptTemplate
+
+from src.core.graph_state import GraphState
+from src.core.embeddings import load_embeddings
+from src.core.llm import load_llm
+from src.vector_store.vector_store import build_vector_store
+from src.config.config import K_OFFSET, MMR_LAMBDA
+from src.exceptions import VectorStoreNotInitializedError, LLMInvocationError
+
+
+class ProjectRAGGraph:
+    def __init__(self):
+        self.embeddings = load_embeddings()
+        self.llm = load_llm()
+        self.vector_store = None
+        self.pdf_count = 0
+        self.memory = MemorySaver()
+        self.workflow = self._build_graph()
+
+    def process_documents(self, pdf_paths, original_names=None):
+        self.pdf_count = len(pdf_paths)
+        self.vector_store = build_vector_store(
+            pdf_paths,
+            self.embeddings,
+            original_names
+        )
+
+    # ---------- Graph Nodes ----------
+
+    def retrieve(self, state: GraphState):
+        if not self.vector_store:
+            raise VectorStoreNotInitializedError("Vector store not initialized")
+
+        k_value = max(1, self.pdf_count + K_OFFSET)
+
+        retriever = self.vector_store.as_retriever(
+            search_type="mmr",
+            search_kwargs={"k": k_value, "lambda_mult": MMR_LAMBDA}
+        )
+
+        documents = retriever.invoke(state["question"])
+        return {"context": documents}
+
+    def generate(self, state: GraphState):
+        try:
+            prompt = ChatPromptTemplate.from_template(
+                """
+                You are an expert Project Analyst.
+                Answer ONLY using the provided context.
+                If the answer is not present, say "I don't know".
+
+                Context:
+                {context}
+
+                Question:
+                {question}
+                """
+            )
+
+            formatted_context = "\n\n".join(
+                doc.page_content for doc in state["context"]
+            )
+
+            chain = prompt | self.llm
+            response = chain.invoke({
+                "context": formatted_context,
+                "question": state["question"]
+            })
+
+            return {"answer": response.content}
+
+        except Exception as e:
+            raise LLMInvocationError(f"LLM failed: {e}")
+
+    # ---------- Graph Build ----------
+
+    def _build_graph(self):
+        workflow = StateGraph(GraphState)
+        workflow.add_node("retrieve", self.retrieve)
+        workflow.add_node("generate", self.generate)
+        workflow.set_entry_point("retrieve")
+        workflow.add_edge("retrieve", "generate")
+        workflow.add_edge("generate", END)
+        return workflow.compile(checkpointer=self.memory)
+
+    def query(self, question: str, thread_id: str):
+        config = {"configurable": {"thread_id": thread_id}}
+        result = self.workflow.invoke({"question": question}, config=config)
+        return result["answer"]

src/vector_store/vector_store.py

@@ -0,0 +1,31 @@
+# src/vector_store.py
+from langchain_community.document_loaders import PyPDFLoader
+from langchain_text_splitters import RecursiveCharacterTextSplitter
+from langchain_community.vectorstores import FAISS
+from src.config.config import CHUNK_SIZE, CHUNK_OVERLAP
+from src.exceptions import DocumentProcessingError
+
+def build_vector_store(pdf_paths, embeddings, original_names=None):
+    try:
+        all_docs = []
+
+        for i, path in enumerate(pdf_paths):
+            loader = PyPDFLoader(path)
+            docs = loader.load()
+
+            if original_names and i < len(original_names):
+                for doc in docs:
+                    doc.metadata["source"] = original_names[i]
+
+            all_docs.extend(docs)
+
+        splitter = RecursiveCharacterTextSplitter(
+            chunk_size=CHUNK_SIZE,
+            chunk_overlap=CHUNK_OVERLAP
+        )
+
+        splits = splitter.split_documents(all_docs)
+        return FAISS.from_documents(splits, embeddings)
+
+    except Exception as e:
+        raise DocumentProcessingError(f"PDF processing failed: {e}")

streamlit_app.py

@@ -0,0 +1,121 @@
+import streamlit as st
+import os
+import tempfile
+# from src.RAG_builder import ProjectRAGGraph # Ensure your graph class is in your_filename.py
+
+from src.rag_graph import ProjectRAGGraph
+
+# from src.graph.rag_graph import ProjectRAGGraph
+# --- Page Config ---
+st.set_page_config(page_title="Project Analyst RAG", layout="wide")
+st.title("📄 Professional Project Analyst Chat")
+
+# --- Initialize Session State ---
+if "rag_graph" not in st.session_state:
+    st.session_state.rag_graph = ProjectRAGGraph()
+if "messages" not in st.session_state:
+    st.session_state.messages = []
+if "thread_id" not in st.session_state:
+    st.session_state.thread_id = "default_user_1" # Hardcoded for demo, could be unique per session
+
+# --- Sidebar: File Upload ---
+with st.sidebar:
+    st.header("Upload Documents")
+    uploaded_files = st.file_uploader(
+        "Upload Project PDFs", 
+        type="pdf", 
+        accept_multiple_files=True
+    )
+    
+    process_button = st.button("Process Documents")
+
+    if process_button and uploaded_files:
+        with st.spinner("Processing PDFs..."):
+            pdf_paths = []
+            original_names = [] # <--- Add this
+            for uploaded_file in uploaded_files:
+                original_names.append(uploaded_file.name) # <--- Capture real name
+                with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp:
+                    tmp.write(uploaded_file.getvalue())
+                    pdf_paths.append(tmp.name)
+            
+            # Pass BOTH the paths and the original names
+            st.session_state.rag_graph.process_documents(
+                pdf_paths, 
+                original_names=original_names
+            )
+            
+            for path in pdf_paths:
+                os.remove(path)
+            st.success("Documents Indexed Successfully!")
+
+    # if process_button and uploaded_files:
+    #     with st.spinner("Processing PDFs..."):
+    #         # Create temporary file paths to pass to your PDF Loader
+    #         pdf_paths = []
+    #         for uploaded_file in uploaded_files:
+    #             with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp:
+    #                 tmp.write(uploaded_file.getvalue())
+    #                 pdf_paths.append(tmp.name)
+            
+    #         # Use your existing process_documents method
+    #         st.session_state.rag_graph.process_documents(pdf_paths)
+            
+    #         # Clean up temp files
+    #         for path in pdf_paths:
+    #             os.remove(path)
+                
+    #         st.success("Documents Indexed Successfully!")
+
+# --- Chat Interface ---
+# Display existing messages
+for message in st.session_state.messages:
+    with st.chat_message(message["role"]):
+        st.markdown(message["content"])
+        if "citations" in message and message["citations"]:
+            with st.expander("View Sources"):
+                for doc in message["citations"]:
+                    st.caption(f"Source: {doc.metadata.get('source', 'Unknown')} - Page: {doc.metadata.get('page', 'N/A')}")
+                    st.write(f"_{doc.page_content[:200]}..._")
+
+# User Input
+if prompt := st.chat_input("Ask a question about your projects..."):
+    # Check if vector store is ready
+    if st.session_state.rag_graph.vector_store is None:
+        st.error("Please upload and process documents first!")
+    else:
+        # Add user message to state
+        st.session_state.messages.append({"role": "user", "content": prompt})
+        with st.chat_message("user"):
+            st.markdown(prompt)
+
+        # Generate Response using the Graph
+        with st.chat_message("assistant"):
+            with st.spinner("Analyzing..."):
+                # We need to call the graph. We'll modify the query return slightly to get citations
+                config = {"configurable": {"thread_id": st.session_state.thread_id}}
+                inputs = {"question": prompt}
+                
+                # Execute graph
+                result = st.session_state.rag_graph.workflow.invoke(inputs, config=config)
+                
+                answer = result["answer"]
+                context = result["context"] # These are the retrieved Document objects
+
+                st.markdown(answer)
+                
+                # Citations section
+                if context:
+                    with st.expander("View Sources"):
+                        for doc in context:
+                            source_name = os.path.basename(doc.metadata.get('source', 'Unknown'))
+                            page_num = doc.metadata.get('page', 0) + 1
+                            st.caption(f"📄 {source_name} (Page {page_num})")
+                            st.write(f"_{doc.page_content[:300]}..._")
+
+                # Add assistant response to state
+                st.session_state.messages.append({
+                    "role": "assistant", 
+                    "content": answer, 
+                    "citations": context
+                })