Update

README.md

@@ -1,8 +1,8 @@
 ---
-title: "GHAPA: Intelligent Health AI Assistant"
-emoji: "🧠⚡"
-colorFrom: "blue"
-colorTo: "green"
+title: "MediAgent AI: Your Trusted AI Health & Drug Navigator"
+emoji: "🧬"
+colorFrom: "indigo"
+colorTo: "purple"
 sdk: "gradio"
 app_file: "app.py"
 pinned: false
@@ -10,36 +10,61 @@ secrets:
   - MODAL_BACKEND_URL
 ---
 
-# 🧠 GHAPA: Global Health AI Prompt Assistant (Powered by Modal ⚡)
+# 🧬 MediAgent AI: Your Trusted AI Health & Drug Navigator
 
-**GHAPA** is a multilingual biomedical assistant that provides trusted, citable health information across global languages. It combines real-time hybrid search with domain-specific LLM reasoning in a modular Retrieval-Augmented Generation (RAG) pipeline.
+**MediAgent AI** is a sophisticated, multilingual health assistant designed to provide reliable, sourced, and easy-to-understand answers to a wide range of medical questions. It leverages a powerful agentic framework to reason, select tools, and synthesize information from trusted sources.
 
-## 🌍 Key Features
+This application is built with a decoupled architecture: a modern Gradio UI hosted on Hugging Face Spaces and a powerful, serverless GPU backend running on [Modal](https://modal.com).
 
-- 🌐 **Multilingual Q&A**: Accepts questions in **English**, **Hindi**, **Spanish**, and **French**, with automatic language detection and translation.
-- 🔍 **Hybrid Search**: Retrieves biomedical data from **PubMed** and general health insights via **DuckDuckGo**, ensuring broader yet reliable context.
-- 📚 **Citable Results**: Every answer includes a direct **source link** and **context snippet**.
-- 🧠 **Biomedical Reasoning**: Uses **PMC-LLaMA**, a large language model fine-tuned on medical literature, for precise and grounded answers.
-- ⚙️ **Serverless Backend**: All heavy processing runs on scalable GPU backends via [Modal](https://modal.com), decoupled from the lightweight Gradio UI.
-- 💡 **Minimal UI**: Clean, responsive interface built using [Gradio](https://www.gradio.app), hosted on [Hugging Face Spaces](https://huggingface.co/spaces).
+## 🚀 Key Features
+
+- 💬 **Multilingual Support**: Automatically detects and responds in **English**, **Hindi**, **Spanish**, and **French**.
+- 🛠️ **Agentic Tool Use**: Instead of a simple search, MediAgent AI actively chooses the best tool for your question from its specialized toolkit:
+    - `search_pubmed`: For technical and clinical research.
+    - `search_web`: For general health and symptom information.
+    - `check_drug_interactions`: To check for interactions between multiple drugs via the openFDA API.
+    - `check_drug_side_effects`: To find adverse reactions for a specific drug via the openFDA API.
+- 🛡️ **Advanced Safety Layers**: Features a critical pre-check system that identifies queries related to **physical medical emergencies** or **mental health crises** and provides immediate, appropriate guidance instead of attempting an answer.
+- 🧠 **Agentic Reasoning Engine**: Powered by **Mistral-7B-Instruct-v0.2** and a **LlamaIndex ReAct Agent**, allowing the model to think, act, and observe in a loop to arrive at a well-reasoned conclusion.
+- 📚 **Sourced & Explained Answers**: All information is presented with citations (links to PubMed/web sources or notes from the FDA database) and complex medical terms are automatically explained in simple language.
+- ✨ **Modern & User-Friendly UI**: A clean, responsive interface built with [Gradio](https://www.gradio.app) that includes example prompts and a one-click "Copy to Clipboard" button for convenience.
 
 ---
 
-## 🛠 Architecture Overview
+## 🛠️ Architecture Overview
+
+MediAgent AI uses an advanced agentic architecture, which is more dynamic than a standard RAG pipeline.
 
 ```plaintext
-         User Input (any language)
-                 ↓
-     Language Detection & Translation
-                 ↓
-     Hybrid Search (PubMed + DuckDuckGo)
-                 ↓
-   Unified Ranking (SentenceTransformer)
-                 ↓
-   Prompt Construction with Top Contexts
-                 ↓
-   PMC-LLaMA Generative Reasoning Engine
-                 ↓
-     Back-Translation & Markdown Output
-                 ↓
-        Final Answer + Cited Sources
+                   User Input (any language)
+                            │
+              ┌─────────────┴─────────────┐
+              │    Gradio UI on HF Spaces │
+              └─────────────┬─────────────┘
+                            │ (API Call to Modal)
+              ┌─────────────↓─────────────┐
+              │  Modal Serverless Backend │
+              └─────────────┬─────────────┘
+                            │
+           1. 🛡️ Critical Safety Pre-Check
+           (Is it a medical or mental health emergency?)
+                            │
+           2. 🌐 Language Detection & Translation to English
+                            │
+           3. 🤖 LlamaIndex ReAct Agent (Mistral-7B)
+                  (Agentic Loop Begins)
+                  a. 🤔 Thought: "Which tool should I use based on the user's query?"
+                  b. 🎬 Action: Selects and executes a tool (e.g., check_drug_interactions)
+                  c. 👀 Observation: Receives data back from the tool (e.g., FDA results)
+                  (Loop continues until the agent has enough information)
+                            │
+           4. 📝 Final Answer Synthesis
+         (Agent composes a complete answer, explaining complex terms)
+                            │
+           5. 🌐 Back-Translation to Original Language
+                            │
+           6. 📤 Formatted Markdown Response
+                            │
+              ┌─────────────↓─────────────┐
+              │  Final Answer Displayed in UI  │
+              └─────────────────────────────┘

app.py

@@ -7,21 +7,19 @@ import uuid
 MODAL_BACKEND_URL = os.getenv("MODAL_BACKEND_URL")
 
 # --- Session ID ---
-SESSION_ID = str(uuid.uuid4())  # Generate a persistent session ID
+SESSION_ID = str(uuid.uuid4())
 
 # --- Backend Communication ---
 def call_modal_backend(user_input):
+    """Sends the user's question to the Modal backend and returns the response."""
     if not MODAL_BACKEND_URL:
         return {
-            "final_markdown": "### ❗ Configuration Error\n**Reason**: `MODAL_BACKEND_URL` is not set.",
-            "detected_lang": "Unknown"
+            "final_markdown": "### ❗ Configuration Error\n**Reason**: `MODAL_BACKEND_URL` environment variable is not set.",
+            "detected_lang": "Error"
         }
 
     headers = {"Content-Type": "application/json"}
-    data = {
-        "user_input": user_input,
-        "session_id": SESSION_ID,
-    }
+    data = {"user_input": user_input, "session_id": SESSION_ID}
 
     try:
         response = requests.post(MODAL_BACKEND_URL, headers=headers, json=data, timeout=300)
@@ -33,52 +31,99 @@ def call_modal_backend(user_input):
             "detected_lang": "Error"
         }
 
+# --- UI Styling ---
+CUSTOM_CSS = """
+body { font-family: 'Segoe UI', Tahoma, Geneva, Verdana, sans-serif; }
+.gradio-container { max-width: 800px !important; margin: auto; padding-top: 2rem; }
+.output-area {
+    background-color: #f9f9f9;
+    padding: 1.5rem;
+    border-radius: 0.5rem;
+    border: 1px solid #e0e0e0;
+    box-shadow: 0 1px 3px rgba(0,0,0,0.05);
+}
+footer { display: none !important; }
+#user-input textarea { font-size: 1rem; line-height: 1.5; }
+"""
+
 # --- Gradio UI Definition ---
-with gr.Blocks(theme=gr.themes.Soft(), css="footer {display: none !important;}") as demo:
+# UPDATED: Title and header to reflect the new name "MediAgent AI"
+with gr.Blocks(theme=gr.themes.Soft(), css=CUSTOM_CSS, title="MediAgent AI") as demo:
+    # --- Header ---
     gr.Markdown("""
-    # 🧠 GHAPA: Global Health AI Prompt Assistant  
-    _Ask trusted health questions in **English**, **Hindi**, **Spanish**, or **French**._
+    # 🧬 MediAgent AI
+    _Your Trusted AI Health & Drug Navigator_
     """)
 
-    with gr.Row():
+    # --- Input Section ---
+    with gr.Group():
         inp = gr.Textbox(
-            label="❓ Enter your health question",
-            placeholder="e.g., What are the symptoms of dengue?",
-            scale=4
+            label="❓ Ask any health question (e.g., about symptoms, drugs, or diseases)",
+            placeholder="e.g., What are the symptoms of dengue? or Do aspirin and warfarin interact?",
+            lines=3,
+            elem_id="user-input"
         )
-        btn = gr.Button("🔍 Submit", variant="primary", scale=1)
+        with gr.Row():
+            btn = gr.Button("🔍 Get Answer", variant="primary", scale=2)
+            clear_btn = gr.ClearButton(value="Clear", scale=1)
+        
+        gr.Examples(
+            examples=[
+                ["What are symptoms of malaria?"],
+                ["Do aspirin and warfarin interact?"],
+                ["¿Cuáles son los efectos secundarios del ibuprofeno?"],
+                ["COVID-19 के लक्षण क्या हैं?"],
+            ],
+            inputs=inp,
+            label="Example Questions"
+        )
+    
+    # --- Output Section ---
+    lang_detected = gr.Label(value="", visible=False)
+    
+    with gr.Column(elem_classes=["output-area"]):
+        output = gr.Markdown(label="✅ Answer & Explanation")
+    
+    copy_btn = gr.Button("📋 Copy Answer to Clipboard", visible=False)
 
-    lang_detected = gr.Markdown("🌐 Language detected: _waiting..._", visible=False)
-    output = gr.Markdown(label="✅ Answer & Explanation")
+    # --- Backend State ---
     backend_result = gr.State()
 
+    # --- UI Logic Functions ---
     def start_processing(user_input):
         if not user_input.strip():
             return {
-                inp: gr.update(),
-                btn: gr.update(),
-                output: "Please enter a question before submitting.",
+                output: gr.Markdown("Please enter a question before submitting."),
                 lang_detected: gr.update(visible=False),
+                copy_btn: gr.update(visible=False),
             }
         return {
             inp: gr.update(interactive=False),
-            btn: gr.update(interactive=False, value="⏳ Processing..."),
+            btn: gr.update(interactive=False, value="⏳ Thinking..."),
             output: "",
-            lang_detected: gr.update(visible=True, value="🌐 Detecting language and processing..."),
+            lang_detected: gr.update(visible=True, value="🌐 Analyzing your question..."),
+            copy_btn: gr.update(visible=False),
         }
 
     def finish_processing(response_json):
+        final_markdown = response_json.get("final_markdown", "Error: No response from backend.")
+        detected_lang = response_json.get('detected_lang', 'Unknown')
+        
+        show_copy = "error" not in detected_lang.lower() and "error" not in final_markdown.lower()
+
         return {
-            inp: gr.update(interactive=True),
-            btn: gr.update(interactive=True, value="🔍 Submit"),
-            output: response_json.get("final_markdown", "Error: No response from backend."),
-            lang_detected: gr.update(visible=True, value=f"🌐 Language detected: **{response_json.get('detected_lang', 'Unknown')}**"),
+            inp: gr.update(interactive=True, value=""),
+            btn: gr.update(interactive=True, value="🔍 Get Answer"),
+            output: final_markdown,
+            lang_detected: gr.update(visible=True, value=f"🌐 Language detected: **{detected_lang}**"),
+            copy_btn: gr.update(visible=show_copy),
         }
-
-    btn.click(
+    
+    # --- Event Listeners ---
+    event_chain = btn.click(
         fn=start_processing,
         inputs=[inp],
-        outputs=[inp, btn, output, lang_detected]
+        outputs=[inp, btn, output, lang_detected, copy_btn]
     ).then(
         fn=call_modal_backend,
         inputs=[inp],
@@ -86,8 +131,20 @@ with gr.Blocks(theme=gr.themes.Soft(), css="footer {display: none !important;}")
     ).then(
         fn=finish_processing,
         inputs=[backend_result],
-        outputs=[inp, btn, output, lang_detected]
+        outputs=[inp, btn, output, lang_detected, copy_btn]
+    )
+
+    clear_btn.add([inp, output, lang_detected, copy_btn])
+
+    copy_btn.click(
+        fn=None,
+        inputs=[output],
+        outputs=None,
+        _js="""(text) => {
+            navigator.clipboard.writeText(text);
+            alert('Answer copied to clipboard!');
+        }"""
     )
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()