add flask files

.gitignore

@@ -0,0 +1,70 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+env/
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+.hypothesis/
+.pytest_cache/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# pyenv
+.python-version
+
+# mypy
+.mypy_cache/
+.dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# dotenv environment variables file
+.env
+
+# VS Code
+.vscode/
+
+# Exclude all PDF files
+*.pdf 

Dockerfile

@@ -0,0 +1,33 @@
+# Use Python 3.11 as the base image
+FROM python:3.11-slim
+
+# Set working directory
+WORKDIR /app
+
+# Install system dependencies
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    && rm -rf /var/lib/apt/lists/*
+
+# Copy requirements first to leverage Docker cache
+COPY requirements.txt .
+
+# Install Python dependencies
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copy the rest of the application
+COPY . .
+
+# Create necessary directories if they don't exist
+RUN mkdir -p private Knowledge_base templates
+
+# Expose the port the app runs on
+EXPOSE 7860
+
+# Set environment variables
+ENV FLASK_APP=app.py
+ENV FLASK_ENV=production
+ENV FLASK_RUN_PORT=7860
+
+# Command to run the application
+CMD ["flask", "run", "--host=0.0.0.0", "--port=7860"] 

Knowledge_base/faiss_index.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:460e6632415b4d967dc7e2e45005fd701fff3cc5516b926d2d9e2329278ed3b1
+size 391725

Knowledge_base/metadata.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16702041923f94f9b6536e771e1b53a8f802966d31709713280c38ee0e7a6aaa
+size 251917

RAG.py

@@ -0,0 +1,122 @@
+import os
+import warnings
+import pickle
+from together import Together
+import faiss
+from sentence_transformers import SentenceTransformer
+from PyPDF2 import PdfReader
+import glob
+
+warnings.filterwarnings("ignore")
+
+# Hardcoded Together API Key
+TOGETHER_API_KEY = "81da53aa3044c7ebead342fb048f016a4e593a86928a783a6fdcc1e3883054e4"
+client = Together(api_key=TOGETHER_API_KEY)
+
+# Initialize embedding model
+embedding_model = SentenceTransformer(
+    "sentence-transformers/all-MiniLM-L6-v2",
+    use_auth_token=os.environ.get("HUGGINGFACE_HUB_TOKEN"),
+)
+
+# Global variables for storing embeddings and documents
+documents = []
+filenames = []
+index = None
+
+def load_index():
+    """Load the FAISS index and document metadata."""
+    global index, documents, filenames
+    
+    if not os.path.exists("knowledge_base/faiss_index.bin") or not os.path.exists("knowledge_base/metadata.pkl"):
+        raise ValueError("Index files not found. Please run preprocess.py first!")
+    
+    print("Loading index and metadata...")
+    index = faiss.read_index("knowledge_base/faiss_index.bin")
+    
+    with open("knowledge_base/metadata.pkl", "rb") as f:
+        metadata = pickle.load(f)
+        documents = metadata["documents"]
+        filenames = metadata["filenames"]
+    
+    print("Index and metadata loaded successfully!")
+
+def answer_question(query: str) -> str:
+    """
+    Answer a question using the RAG system.
+    
+    Args:
+        query (str): The user's question
+        
+    Returns:
+        str: The generated answer
+    """
+    global index
+    
+    # Load index if not already loaded
+    if index is None:
+        load_index()
+    
+    # Get query embedding
+    query_embedding = embedding_model.encode([query])
+    faiss.normalize_L2(query_embedding)
+    
+    # Get top similar documents
+    scores, indices = index.search(query_embedding, min(3, len(documents)))
+    
+    # Build context from retrieved documents
+    context_parts = []
+    relevant_docs = []
+    
+    for score, idx in zip(scores[0], indices[0]):
+        if idx < len(documents):
+            doc_info = {
+                "content": documents[idx],
+                "filename": filenames[idx],
+                "score": float(score),
+            }
+            relevant_docs.append(doc_info)
+            context_parts.append(f"[{doc_info['filename']}]\n{doc_info['content']}")
+    
+    if not relevant_docs:
+        return "I couldn't find any relevant information to answer your question."
+    
+    # Combine context
+    context = "\n\n".join(context_parts)
+    
+    # Create prompt for LLM
+    llm_prompt = f"""Answer the question based on the provided context documents.
+
+Context:
+{context}
+
+Question: {query}
+
+Instructions:
+- Answer based only on the information in the context
+- If the context doesn't contain enough information, say so
+- Mention which document(s) you're referencing
+- Start with According to [document name]
+- Add brackets to the document name
+
+Answer:"""
+    
+    try:
+        # Generate answer using Together AI
+        response = client.chat.completions.create(
+            model="meta-llama/Llama-3.3-70B-Instruct-Turbo",
+            messages=[{"role": "user", "content": llm_prompt}],
+            max_tokens=500,
+            temperature=0.7,
+        )
+        answer = response.choices[0].message.content
+        
+        # Add source information
+        sources_list = [doc["filename"] for doc in relevant_docs]
+        sources_text = sources_list[0]
+        full_answer = f"{answer}\n\n📄 Source Used: {sources_text}"
+        
+        return full_answer
+        
+    except Exception as e:
+        return f"Error generating answer: {str(e)}"

app.py

@@ -0,0 +1,85 @@
+from flask import Flask, render_template, request, jsonify
+from dotenv import load_dotenv
+import os
+import json
+import httpx
+
+# Explicitly load the .env file from the script's directory
+#dotenv_path = os.path.join(os.path.dirname(__file__), '.env')
+#load_dotenv(dotenv_path=dotenv_path)
+
+from RAG import answer_question
+
+app = Flask(__name__)
+
+# Load Surge configuration
+try:
+    with open("private/surge_config.json", "r") as f:
+        surge_config = json.load(f)
+except FileNotFoundError:
+    print("Warning: surge_config.json not found. SMS functionality will be disabled.")
+    surge_config = None
+
+@app.route('/')
+def index():
+    return render_template('index.html')
+
+@app.route('/chat', methods=['POST'])
+def chat():
+    try:
+        data = request.get_json()
+        user_question = data.get('question', '')
+        
+        if not user_question:
+            return jsonify({'error': 'No question provided'}), 400
+            
+        # Get answer from RAG system
+        answer = answer_question(user_question)
+        
+        return jsonify({
+            'answer': answer
+        })
+        
+    except Exception as e:
+        return jsonify({'error': str(e)}), 500
+
+@app.route('/send-sms', methods=['POST'])
+def send_sms():
+    if not surge_config:
+        return jsonify({'error': 'SMS functionality not configured'}), 503
+        
+    try:
+        data = request.get_json()
+        message_content = data.get('message', '')
+        
+        if not message_content:
+            return jsonify({'error': 'No message provided'}), 400
+            
+        # Send SMS via Surge API
+        with httpx.Client() as client:
+            response = client.post(
+                "https://api.surgemsg.com/messages",
+                headers={
+                    "Authorization": f"Bearer {surge_config['api_key']}",
+                    "Surge-Account": surge_config["account_id"],
+                    "Content-Type": "application/json",
+                },
+                json={
+                    "body": message_content,
+                    "conversation": {
+                        "contact": {
+                            "first_name": surge_config["my_first_name"],
+                            "last_name": surge_config["my_last_name"],
+                            "phone_number": surge_config["my_phone_number"],
+                        }
+                    },
+                },
+            )
+            response.raise_for_status()
+            return jsonify({'status': 'Message sent successfully'})
+            
+    except Exception as e:
+        return jsonify({'error': str(e)}), 500
+
+if __name__ == '__main__':
+    app.run(debug=True)

preprocess.py

@@ -0,0 +1,96 @@
+import os
+import warnings
+import pickle
+import faiss
+from sentence_transformers import SentenceTransformer
+from PyPDF2 import PdfReader
+import glob
+from together import Together
+
+warnings.filterwarnings("ignore")
+
+# Hardcoded Together API Key
+TOGETHER_API_KEY = "81da53aa3044c7ebead342fb048f016a4e593a86928a783a6fdcc1e3883054e4"
+client = Together(api_key=TOGETHER_API_KEY)
+
+# Initialize embedding model
+embedding_model = SentenceTransformer(
+    "sentence-transformers/all-MiniLM-L6-v2",
+    use_auth_token=os.environ.get("HUGGINGFACE_HUB_TOKEN"),
+)
+
+def extract_text_from_pdf(pdf_path):
+    """Extract text from a PDF file."""
+    try:
+        reader = PdfReader(pdf_path)
+        text = ""
+        for page in reader.pages:
+            text += page.extract_text() + "\n"
+        return text.strip()
+    except Exception as e:
+        print(f"Error processing {pdf_path}: {str(e)}")
+        return ""
+
+def create_index():
+    """Create and save the FAISS index and document metadata."""
+    # Create knowledge_base directory if it doesn't exist
+    os.makedirs("knowledge_base", exist_ok=True)
+    
+    # Get all PDF files from knowledge_base directory
+    pdf_files = glob.glob("Knowledge_base/*.pdf")
+    
+    if not pdf_files:
+        raise ValueError("No PDF files found in Knowledge_base directory!")
+    
+    print(f"Found {len(pdf_files)} PDF files. Processing...")
+    
+    # Process documents
+    documents = []
+    filenames = []
+    
+    for pdf_path in pdf_files:
+        filename = os.path.basename(pdf_path)
+        content = extract_text_from_pdf(pdf_path)
+        
+        if content:
+            # Split content into chunks (roughly 1000 characters each)
+            chunks = [content[i:i+1000] for i in range(0, len(content), 1000)]
+            
+            for i, chunk in enumerate(chunks):
+                if chunk.strip():
+                    documents.append(chunk)
+                    filenames.append(f"{filename} (chunk {i+1})")
+    
+    if not documents:
+        raise ValueError("No valid content extracted from PDFs!")
+    
+    print(f"Successfully processed {len(documents)} chunks from {len(pdf_files)} PDFs")
+    
+    # Create embeddings
+    print("Creating embeddings...")
+    embeddings = embedding_model.encode(documents)
+    
+    # Set up FAISS index
+    dimension = embeddings.shape[1]
+    index = faiss.IndexFlatIP(dimension)
+    
+    # Normalize embeddings for cosine similarity
+    faiss.normalize_L2(embeddings)
+    index.add(embeddings)
+    
+    # Save the index and metadata
+    print("Saving index and metadata...")
+    faiss.write_index(index, "knowledge_base/faiss_index.bin")
+    
+    metadata = {
+        "documents": documents,
+        "filenames": filenames
+    }
+    
+    with open("knowledge_base/metadata.pkl", "wb") as f:
+        pickle.dump(metadata, f)
+    
+    print("Index and metadata saved successfully!")
+
+if __name__ == "__main__":
+    create_index() 

private/surge_config.json

@@ -0,0 +1,7 @@
+{
+    "api_key": "sk_live_b64524jqzpp4koyu2lzgzlhdaglxdppidj2azpwerqjnxkphpmn4r6wb",
+    "account_id": "acct_01jxrvt3v0f8t8qs1aewyx021m",
+    "my_phone_number": "6044467277",
+    "my_first_name": "Lei Lei",
+    "my_last_name": "Wu"
+}

requirements.txt

@@ -0,0 +1,7 @@
+flask==3.0.2
+python-dotenv==1.0.1
+httpx==0.27.0
+together==0.2.5
+faiss-cpu==1.7.4
+sentence-transformers==2.5.1
+PyPDF2==3.0.1 

templates/index.html

@@ -0,0 +1,211 @@
+<!DOCTYPE html>
+<html lang="en">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>Kidney Key Chat</title>
+    <link href="https://cdn.jsdelivr.net/npm/[email protected]/dist/css/bootstrap.min.css" rel="stylesheet">
+    <style>
+        body {
+            background: linear-gradient(135deg, #e0eafc 0%, #cfdef3 100%);
+            min-height: 100vh;
+        }
+        .app-container {
+            max-width: 700px;
+            margin: 40px auto;
+            background: #fff;
+            border-radius: 18px;
+            box-shadow: 0 8px 32px 0 rgba(31, 38, 135, 0.18);
+            padding: 32px 28px 24px 28px;
+        }
+        .chat-container {
+            height: 400px;
+            overflow-y: auto;
+            border: 1.5px solid #b6c6e0;
+            padding: 24px 18px;
+            margin-bottom: 24px;
+            border-radius: 12px;
+            background: #f8fbff;
+        }
+        .message {
+            margin-bottom: 18px;
+            padding: 14px 18px;
+            border-radius: 12px;
+            font-size: 1.08rem;
+            line-height: 1.5;
+            max-width: 80%;
+            word-break: break-word;
+            box-shadow: 0 2px 8px rgba(80, 120, 200, 0.07);
+        }
+        .user-message {
+            background: linear-gradient(90deg, #a1c4fd 0%, #c2e9fb 100%);
+            color: #1a237e;
+            margin-left: auto;
+            text-align: right;
+        }
+        .bot-message {
+            background: linear-gradient(90deg, #fbc2eb 0%, #a6c1ee 100%);
+            color: #4a148c;
+            margin-right: auto;
+            text-align: left;
+        }
+        .loading {
+            display: none;
+            text-align: center;
+            margin: 10px 0;
+        }
+        .sms-button {
+            margin-top: 10px;
+        }
+        .app-title {
+            font-family: 'Segoe UI', 'Roboto', Arial, sans-serif;
+            font-weight: 700;
+            font-size: 2.3rem;
+            color: #283593;
+            letter-spacing: 1px;
+        }
+        .app-desc {
+            font-size: 1.08rem;
+            color: #374151;
+            margin-bottom: 28px;
+            margin-top: 8px;
+            background: #e3f0ff;
+            border-radius: 8px;
+            padding: 12px 18px;
+        }
+        .input-group input {
+            font-size: 1.08rem;
+            border-radius: 8px 0 0 8px;
+        }
+        .input-group .btn-primary {
+            border-radius: 0 8px 8px 0;
+            font-weight: 600;
+            background: linear-gradient(90deg, #43cea2 0%, #185a9d 100%);
+            border: none;
+        }
+        .input-group .btn-primary:hover {
+            background: linear-gradient(90deg, #185a9d 0%, #43cea2 100%);
+        }
+        .btn-success.sms-button {
+            background: linear-gradient(90deg, #f7971e 0%, #ffd200 100%);
+            color: #333;
+            border: none;
+            font-weight: 600;
+        }
+        .btn-success.sms-button:hover {
+            background: linear-gradient(90deg, #ffd200 0%, #f7971e 100%);
+            color: #222;
+        }
+        @media (max-width: 600px) {
+            .app-container {
+                padding: 10px 2px 18px 2px;
+            }
+            .chat-container {
+                padding: 10px 4px;
+            }
+        }
+    </style>
+</head>
+<body>
+    <div class="app-container">
+        <h1 class="text-center app-title mb-2">Kidney Key Chat</h1>
+        <div class="app-desc text-center">
+            <strong>What is this?</strong> <br>
+            Kidney Key Chat is an AI-powered assistant designed to help healthcare professionals and patients with questions about renal dosing, dialysis, and medication adjustments for kidney function. <br><br>
+            <strong>How does it work?</strong> <br>
+            Just type your question below (e.g., "How do you dose amoxicillin in CKD?"). The app uses Retrieval-Augmented Generation (RAG) to search a curated knowledge base of clinical PDFs, finds the most relevant information, and generates a clear, referenced answer. You can also send the chatbot's response to your phone via SMS for easy reference.
+        </div>
+        <div class="chat-container" id="chatContainer">
+            <!-- Messages will be added here -->
+        </div>
+        <div class="loading" id="loading">
+            <div class="spinner-border text-primary" role="status">
+                <span class="visually-hidden">Loading...</span>
+            </div>
+        </div>
+        <div class="input-group mb-3">
+            <input type="text" id="userInput" class="form-control" placeholder="Type your question here...">
+            <button class="btn btn-primary" onclick="sendMessage()">Send</button>
+        </div>
+        <div class="text-center">
+            <button class="btn btn-success sms-button" onclick="sendLastResponse()" id="smsButton" style="display: none;">
+                Send Last Response via SMS
+            </button>
+        </div>
+    </div>
+    <script>
+        let lastBotResponse = '';
+        function addMessage(message, isUser) {
+            const chatContainer = document.getElementById('chatContainer');
+            const messageDiv = document.createElement('div');
+            messageDiv.className = `message ${isUser ? 'user-message' : 'bot-message'}`;
+            messageDiv.textContent = message;
+            chatContainer.appendChild(messageDiv);
+            chatContainer.scrollTop = chatContainer.scrollHeight;
+        }
+        function showLoading() {
+            document.getElementById('loading').style.display = 'block';
+        }
+        function hideLoading() {
+            document.getElementById('loading').style.display = 'none';
+        }
+        async function sendMessage() {
+            const userInput = document.getElementById('userInput');
+            const message = userInput.value.trim();
+            if (!message) return;
+            addMessage(message, true);
+            userInput.value = '';
+            showLoading();
+            try {
+                const response = await fetch('/chat', {
+                    method: 'POST',
+                    headers: {
+                        'Content-Type': 'application/json',
+                    },
+                    body: JSON.stringify({ question: message }),
+                });
+                const data = await response.json();
+                if (response.ok) {
+                    lastBotResponse = data.answer;
+                    addMessage(data.answer, false);
+                    document.getElementById('smsButton').style.display = 'inline-block';
+                } else {
+                    addMessage('Error: ' + data.error, false);
+                }
+            } catch (error) {
+                addMessage('Error: Could not connect to the server', false);
+            } finally {
+                hideLoading();
+            }
+        }
+        async function sendLastResponse() {
+            if (!lastBotResponse) return;
+            showLoading();
+            try {
+                const response = await fetch('/send-sms', {
+                    method: 'POST',
+                    headers: {
+                        'Content-Type': 'application/json',
+                    },
+                    body: JSON.stringify({ message: lastBotResponse }),
+                });
+                const data = await response.json();
+                if (response.ok) {
+                    alert('Message sent successfully!');
+                } else {
+                    alert('Error: ' + data.error);
+                }
+            } catch (error) {
+                alert('Error: Could not send SMS');
+            } finally {
+                hideLoading();
+            }
+        }
+        document.getElementById('userInput').addEventListener('keypress', function(e) {
+            if (e.key === 'Enter') {
+                sendMessage();
+            }
+        });
+    </script>
+</body>
+</html>