Create app2.py

app2.py

@@ -0,0 +1,506 @@
+# If needed in Colab, install first:
+# !pip install -U gradio pinecone llama-index llama-index-vector-stores-pinecone llama-index-readers-file pypdf
+
+from llama_index.core import VectorStoreIndex, SimpleDirectoryReader, StorageContext, Settings
+
+# --- Imports ---
+import logging
+import sys
+import gradio as gr
+import os
+
+from pinecone import Pinecone, ServerlessSpec
+from llama_index.core import VectorStoreIndex, SimpleDirectoryReader, StorageContext, Settings
+from llama_index.vector_stores.pinecone import PineconeVectorStore
+from llama_index.readers.file import PDFReader
+from llama_index.llms.openai import OpenAI
+from llama_index.embeddings.openai import OpenAIEmbedding
+
+
+# --- Logging ---
+logging.basicConfig(stream=sys.stdout, level=logging.INFO)
+
+
+Settings.llm = OpenAI(model="gpt-4o-mini", temperature=0.2)
+Settings.embed_model = OpenAIEmbedding(model="text-embedding-ada-002")
+Settings.chunk_size = 600
+Settings.chunk_overlap = 200
+
+
+# Define a system prompt
+system_prompt = '''
+You are AYesha, the Decoding Data Science (DDS) Enterprise HR Chatbot. Answer questions exclusively using the attached DDS HR Handbook. Base all responses on the most up-to-date information available in the handbook. Only respond to queries directly related to DDS HR policies as outlined in the handbook.
+
+- If a question pertains to topics outside DDS HR policies, respond politely, clarifying that you are a human resources bot and only answer DDS HR questions.
+- For questions you cannot answer (e.g., requests for old policies, salary details, or confidential information), politely decline and direct the user to email connect@decodingdatascience.com.
+- Never answer questions about anything outside of your scope.
+- Persist in following these constraints for any follow-up questions.
+- Before answering, carefully check that the information and query are within the allowed scope. Follow chain-of-thought reasoning:
+  1. First, reason step-by-step whether the question is covered in the current handbook and is within HR.
+  2. Only after confirming, produce a final answer.
+
+Format answers as concise, professional responses. Do not wrap answers in code blocks or any special formatting.
+
+Output requirements:
+- For allowed HR questions, answer concisely based only on the latest DDS HR handbook information.
+- For forbidden topics, output: “I’m sorry, I can only answer questions about the latest DDS HR policies. For confidential or other queries, please email connect@decodingdatascience.com.”
+
+
+**Example 1**
+User: What is the leave encashment policy at DDS?
+Reasoning: This is an HR policy question found in the latest handbook.
+Final Answer: [Provide answer summarized from the latest handbook’s section on leave encashment]
+
+**Example 2**
+User: Can you tell me the salary range for Data Scientists?
+Reasoning: Salary details are confidential and not shared by this bot.
+Final Answer: I’m sorry, I can only answer questions about the latest DDS HR policies. For confidential or other queries, please email connect@decodingdatascience.com.
+
+**Example 3**
+User: Can you explain what DDS does as a company overall?
+Reasoning: This is not an HR question, so it cannot be answered.
+Final Answer: I’m sorry, I only answer DDS HR policy questions as outlined in the handbook.
+
+(Real-world examples should be longer and use precise wording from the handbook where appropriate.)
+
+**Important instructions:**
+- Only answer questions directly supported by the latest DDS HR handbook.
+- Decline politely and redirect to the provided email address for any questions outside scope or for confidential information.
+- Always reason before concluding. Only present the answer after checking scope and source.
+
+Remember: As AYesha, the DDS HR Enterprise Chatbot, you must never provide information outside authorized HR handbook content and always respond respectfully according to these constraints.
+
+'''
+
+
+# --- Load API Key from Hugging face environment ---
+OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
+PINECONE_API_KEY = os.getenv("PINECONE_API_KEY")
+
+
+# --- Initialize Pinecone ---
+pc = Pinecone(api_key=PINECONE_API_KEY)
+index_name = "quickstart"
+dimension = 1536
+
+
+# --- Delete index if it already exists (optional) ---
+existing_indexes = [idx["name"] for idx in pc.list_indexes()]
+
+if index_name in existing_indexes:
+    pc.delete_index(index_name)
+
+
+# --- Create Pinecone index ---
+pc.create_index(
+    name=index_name,
+    dimension=dimension,
+    metric="euclidean",
+    spec=ServerlessSpec(cloud="aws", region="us-east-1"),
+)
+
+pinecone_index = pc.Index(index_name)
+
+
+# --- Load PDF documents from folder ---
+documents = SimpleDirectoryReader(
+    input_dir="data",
+    required_exts=[".pdf"],
+    file_extractor={".pdf": PDFReader()}
+).load_data()
+
+if not documents:
+    raise ValueError("No PDF documents were loaded from the 'data' folder.")
+
+
+# --- Create Vector Index ---
+vector_store = PineconeVectorStore(pinecone_index=pinecone_index)
+storage_context = StorageContext.from_defaults(vector_store=vector_store)
+
+index = VectorStoreIndex.from_documents(
+    documents,
+    storage_context=storage_context
+)
+
+
+# --- Query Engine ---
+query_engine = index.as_query_engine(system_prompt=system_prompt)
+
+
+# --- Gradio App ---
+def query_doc(prompt):
+    try:
+        response = query_engine.query(prompt)
+        return str(response)
+    except Exception as e:
+        return f"Error: {str(e)}"
+
+
+# -------------------------------------------------------------------
+# Professional Gradio UI
+# Only the Gradio interface is updated below.
+# No RAG logic, LLM, embeddings, Pinecone, PDF loading, or prompt rules changed.
+# -------------------------------------------------------------------
+
+CUSTOM_CSS = """
+.gradio-container {
+    max-width: 1180px !important;
+    margin: 0 auto !important;
+}
+
+.dds-hero {
+    border: 1px solid var(--border-color-primary);
+    background: var(--block-background-fill);
+    border-radius: 22px;
+    padding: 28px;
+    margin-bottom: 18px;
+}
+
+.dds-title {
+    font-size: 2rem;
+    font-weight: 750;
+    letter-spacing: -0.02em;
+    margin-bottom: 8px;
+}
+
+.dds-subtitle {
+    font-size: 1rem;
+    color: var(--body-text-color-subdued);
+    max-width: 880px;
+    line-height: 1.6;
+}
+
+.dds-badges {
+    display: flex;
+    flex-wrap: wrap;
+    gap: 8px;
+    margin-top: 18px;
+}
+
+.dds-badge {
+    border: 1px solid var(--border-color-primary);
+    background: var(--background-fill-secondary);
+    color: var(--body-text-color);
+    border-radius: 999px;
+    padding: 7px 12px;
+    font-size: 0.86rem;
+}
+
+.dds-card {
+    border: 1px solid var(--border-color-primary);
+    background: var(--block-background-fill);
+    border-radius: 18px;
+    padding: 18px;
+    margin-bottom: 12px;
+}
+
+.dds-muted {
+    color: var(--body-text-color-subdued);
+    font-size: 0.92rem;
+    line-height: 1.55;
+}
+
+.dds-small-heading {
+    font-size: 1rem;
+    font-weight: 700;
+    margin-bottom: 8px;
+}
+
+.dds-footer {
+    text-align: center;
+    color: var(--body-text-color-subdued);
+    font-size: 0.86rem;
+    margin-top: 16px;
+}
+
+textarea {
+    border-radius: 14px !important;
+}
+
+button {
+    border-radius: 12px !important;
+}
+"""
+
+
+example_questions = [
+    "What is the leave policy at DDS?",
+    "How can I apply for annual leave?",
+    "What should I do if I have an HR-related concern?",
+    "Can you explain the employee code of conduct?",
+    "What is the process for reporting a workplace issue?"
+]
+
+
+def respond(message, history):
+    """
+    UI wrapper only.
+    Calls the existing query_doc() function without changing backend logic.
+    """
+    if history is None:
+        history = []
+
+    message = (message or "").strip()
+
+    if not message:
+        return history, ""
+
+    answer = query_doc(message)
+
+    history = history + [
+        {"role": "user", "content": message},
+        {"role": "assistant", "content": answer}
+    ]
+
+    return history, ""
+
+
+theme = gr.themes.Default(
+    primary_hue="slate",
+    secondary_hue="gray",
+    neutral_hue="gray",
+    spacing_size="md",
+    radius_size="lg",
+    text_size="md"
+)
+
+
+with gr.Blocks(
+    theme=theme,
+    css=CUSTOM_CSS,
+    title="DDS HR Enterprise Chatbot",
+    fill_width=True
+) as demo:
+
+    gr.HTML("""
+    <div class="dds-hero">
+        <div class="dds-title">DDS HR Enterprise Chatbot</div>
+        <div class="dds-subtitle">
+            A professional HR assistant for Decoding Data Science employees.
+            Ask questions related to DDS HR policies, employee guidelines, workplace processes,
+            and handbook-supported information.
+        </div>
+        <div class="dds-badges">
+            <span class="dds-badge">DDS HR Handbook</span>
+            <span class="dds-badge">Policy-Grounded Answers</span>
+            <span class="dds-badge">Professional HR Support</span>
+            <span class="dds-badge">Light & Dark Mode Friendly</span>
+        </div>
+    </div>
+    """)
+
+    with gr.Tabs():
+
+        # -------------------------------------------------------------
+        # Layout 1: Conversational Chat Layout
+        # -------------------------------------------------------------
+        with gr.Tab("HR Chat Assistant"):
+
+            with gr.Row():
+                with gr.Column(scale=3):
+
+                    chatbot = gr.Chatbot(
+                        label="DDS HR Assistant",
+                        height=520,
+                        layout="panel",
+                        placeholder=(
+                            "Ask a DDS HR policy question to get started. "
+                            "Example: What is the leave policy at DDS?"
+                        ),
+                        render_markdown=True,
+                        sanitize_html=True,
+                        buttons=["copy", "copy_all"]
+                    )
+
+                    user_message = gr.Textbox(
+                        label="Your HR Question",
+                        placeholder="Type your DDS HR policy question here...",
+                        lines=3
+                    )
+
+                    with gr.Row():
+                        submit_btn = gr.Button("Ask DDS HR", variant="primary")
+                        clear_btn = gr.ClearButton(
+                            components=[chatbot, user_message],
+                            value="Clear Chat"
+                        )
+
+                with gr.Column(scale=1):
+
+                    gr.HTML("""
+                    <div class="dds-card">
+                        <div class="dds-small-heading">How to use this assistant</div>
+                        <div class="dds-muted">
+                            Ask questions that are directly related to DDS HR policies.
+                            The assistant answers using the approved HR handbook context.
+                            For confidential or unsupported topics, it redirects users to the DDS HR contact email.
+                        </div>
+                    </div>
+                    """)
+
+                    gr.Markdown("### Quick HR Questions")
+
+                    for question in example_questions:
+                        example_btn = gr.Button(question)
+                        example_btn.click(
+                            fn=lambda q=question: q,
+                            inputs=[],
+                            outputs=user_message,
+                            queue=False
+                        )
+
+            submit_btn.click(
+                fn=respond,
+                inputs=[user_message, chatbot],
+                outputs=[chatbot, user_message],
+                show_progress="minimal"
+            )
+
+            user_message.submit(
+                fn=respond,
+                inputs=[user_message, chatbot],
+                outputs=[chatbot, user_message],
+                show_progress="minimal"
+            )
+
+        # -------------------------------------------------------------
+        # Layout 2: Classic Single Q&A Layout
+        # -------------------------------------------------------------
+        with gr.Tab("Classic Q&A View"):
+
+            with gr.Row():
+
+                with gr.Column(scale=1):
+
+                    gr.HTML("""
+                    <div class="dds-card">
+                        <div class="dds-small-heading">Single Question Mode</div>
+                        <div class="dds-muted">
+                            Use this layout when you want a simple question-and-answer experience.
+                            This mode calls the same query_doc() backend function used in the original app.
+                        </div>
+                    </div>
+                    """)
+
+                    single_question = gr.Textbox(
+                        label="Ask a DDS HR Question",
+                        placeholder="Example: What is the leave policy at DDS?",
+                        lines=5
+                    )
+
+                    with gr.Row():
+                        single_submit = gr.Button("Get Answer", variant="primary")
+                        single_clear = gr.ClearButton(
+                            components=[single_question],
+                            value="Clear Question"
+                        )
+
+                    gr.Examples(
+                        examples=[[q] for q in example_questions],
+                        inputs=single_question,
+                        label="Try sample HR questions"
+                    )
+
+                with gr.Column(scale=1):
+
+                    single_answer = gr.Textbox(
+                        label="DDS HR Answer",
+                        lines=16,
+                        buttons=["copy"]
+                    )
+
+            single_submit.click(
+                fn=query_doc,
+                inputs=single_question,
+                outputs=single_answer,
+                show_progress="minimal"
+            )
+
+            single_question.submit(
+                fn=query_doc,
+                inputs=single_question,
+                outputs=single_answer,
+                show_progress="minimal"
+            )
+
+        # -------------------------------------------------------------
+        # FAQ and Scope Section
+        # -------------------------------------------------------------
+        with gr.Tab("FAQs & Scope"):
+
+            gr.Markdown("## Frequently Asked Questions")
+
+            with gr.Accordion("What can this chatbot answer?", open=True):
+                gr.Markdown(
+                    """
+                    This chatbot can answer questions related to DDS HR policies and handbook-supported employee information.
+
+                    Examples include:
+                    - Leave policy
+                    - Workplace conduct
+                    - HR procedures
+                    - Employee policy guidance
+                    - Handbook-supported HR processes
+                    """
+                )
+
+            with gr.Accordion("Can it answer salary or confidential employee questions?", open=False):
+                gr.Markdown(
+                    """
+                    No. Salary details, confidential employee records, private HR decisions,
+                    and unsupported internal information should not be answered by this assistant.
+
+                    For such questions, employees should email:
+
+                    **connect@decodingdatascience.com**
+                    """
+                )
+
+            with gr.Accordion("Can it answer general questions about DDS as a company?", open=False):
+                gr.Markdown(
+                    """
+                    No. This assistant is scoped specifically to DDS HR policy questions.
+
+                    General business, marketing, training, sales, or company overview questions are outside its HR scope.
+                    """
+                )
+
+            with gr.Accordion("What happens if the answer is not in the handbook?", open=False):
+                gr.Markdown(
+                    """
+                    The assistant should clearly say that it cannot answer based on the available HR handbook content
+                    and redirect the user to:
+
+                    **connect@decodingdatascience.com**
+                    """
+                )
+
+            with gr.Accordion("Is this suitable for light mode and dark mode?", open=False):
+                gr.Markdown(
+                    """
+                    Yes. The interface uses Gradio theme variables and neutral styling so it works cleanly
+                    with both light and dark display modes.
+                    """
+                )
+
+            gr.HTML("""
+            <div class="dds-card">
+                <div class="dds-small-heading">Responsible Use Notice</div>
+                <div class="dds-muted">
+                    This chatbot is designed to support HR policy understanding.
+                    It should not be used as a replacement for official HR decisions,
+                    confidential HR discussions, legal advice, or direct communication with DDS HR.
+                </div>
+            </div>
+            """)
+
+    gr.HTML("""
+    <div class="dds-footer">
+        DDS HR Enterprise Chatbot · Powered by LlamaIndex, Pinecone, OpenAI, and Gradio
+    </div>
+    """)
+
+
+# For Hugging Face Spaces, demo.launch() is enough.
+# If you test locally or in Colab and want a public temporary link, you can use: demo.launch(share=True)
+demo.launch()