Create app.py

app.py

@@ -0,0 +1,64 @@
+import os
+import gradio as gr
+from transformers import pipeline
+
+# Retrieve Groq API key from environment variable
+GROQ_API_KEY = os.getenv("GROQ_API_KEY")
+
+# Local fallback pipelines (replace with Groq API calls as desired)
+classifier = pipeline("zero-shot-classification", model="facebook/bart-large-mnli")
+generator = pipeline("text-generation", model="gpt2", max_length=150)
+summarizer = pipeline("summarization")
+
+def email_assistant(email_text):
+    # Example: If using Groq API, replace below local classifier code with API call using GROQ_API_KEY
+
+    candidate_labels = ["Sales", "Support", "Spam", "Finance", "General"]
+    classification = classifier(email_text, candidate_labels)
+    category = classification['labels'][0]
+    confidence = round(classification['scores'][0]*100, 2)
+
+    priority = "High" if category in ["Sales", "Support"] and confidence > 80 else "Normal"
+
+    forwarding_emails = {
+        "Sales": "sales@company.com",
+        "Support": "support@company.com",
+        "Finance": "finance@company.com",
+        "Spam": None,
+        "General": "info@company.com"
+    }
+    suggested_action = f"Forward to {forwarding_emails[category]}" if forwarding_emails[category] else "No action suggested"
+
+    # Example: For Groq API generation call, replace below with API call using GROQ_API_KEY
+    response_prompt = f"Write a professional reply to this email:\n\n{email_text}\n\nResponse:"
+    response = generator(response_prompt, max_length=150, num_return_sequences=1)[0]['generated_text']
+
+    # Example: For Groq API summarization call, replace below with API call using GROQ_API_KEY
+    action_items = summarizer(email_text, max_length=50, min_length=10, do_sample=False)[0]['summary_text']
+
+    return {
+        "Category": category,
+        "Confidence": f"{confidence}%",
+        "Priority": priority,
+        "Suggested Action": suggested_action,
+        "Draft Response": response,
+        "Action Items": action_items
+    }
+
+with gr.Blocks() as demo:
+    gr.Markdown("# Email Classification & Response Assistant")
+    email_input = gr.TextArea(label="Paste your email here")
+    submit = gr.Button("Analyze Email")
+    output_category = gr.Textbox(label="Category")
+    output_confidence = gr.Textbox(label="Confidence")
+    output_priority = gr.Textbox(label="Priority")
+    output_action = gr.Textbox(label="Suggested Action")
+    output_response = gr.Textbox(label="Draft Response")
+    output_items = gr.Textbox(label="Extracted Action Items")
+    submit.click(email_assistant, inputs=email_input, outputs=[
+        output_category, output_confidence, output_priority,
+        output_action, output_response, output_items
+    ])
+
+if __name__ == "__main__":
+    demo.launch()