Deploy to HF Spaces - Gradio app with Azure Container Apps support

.dockerignore

@@ -0,0 +1,74 @@
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# Virtual environments
+venv/
+ENV/
+env/
+.venv/
+
+# IDE
+.idea/
+.vscode/
+*.swp
+*.swo
+*~
+
+# Git
+.git/
+.gitignore
+
+# Environment files (secrets)
+.env
+.env.local
+.env.*.local
+
+# Testing
+.pytest_cache/
+.coverage
+htmlcov/
+.tox/
+
+# Documentation
+docs/
+*.md
+!README.md
+
+# Logs
+*.log
+logs/
+
+# Temporary files
+tmp/
+temp/
+*.tmp
+
+# OS files
+.DS_Store
+Thumbs.db
+
+# Deployment files (not needed in container)
+.github/
+*.yml
+!requirements.txt
+azure-deploy.sh
+deploy.ps1

DEPLOYMENT_INSTRUCTIONS.md

@@ -0,0 +1,257 @@
+# SAP Finance Dashboard - Azure Container Apps Deployment Guide
+
+## Prerequisites
+
+1. **Azure CLI** - Install from: https://aka.ms/installazurecliwindows
+2. **Azure Subscription** - Active Azure account
+3. **Git** (optional) - For code updates
+
+## Quick Deploy Steps
+
+### Option 1: Automated Deployment Script
+
+Open PowerShell as Administrator and run:
+
+```powershell
+cd "c:\Users\amlal\Downloads\VSCode-SAP-AI-Copilot-Projects2025\SAP-RPT-1-OSS-App"
+.\deploy-azure.ps1
+```
+
+The script will:
+- ✅ Create Azure Resource Group
+- ✅ Create Azure Container Registry
+- ✅ Build Docker image in Azure (no local Docker needed)
+- ✅ Create Container Apps Environment
+- ✅ Deploy your application
+- ✅ Provide public URL
+
+**Deployment Time:** ~10-15 minutes
+
+### Option 2: Manual Deployment (Step-by-Step)
+
+If the automated script doesn't work, follow these manual commands:
+
+#### 1. Login to Azure
+```bash
+az login
+```
+
+#### 2. Set Variables
+```bash
+set RESOURCE_GROUP=rg-sap-finance-dashboard
+set LOCATION=eastus
+set ACR_NAME=acrsapfinance%RANDOM%
+set ENVIRONMENT_NAME=env-sap-finance
+set APP_NAME=sap-finance-dashboard
+```
+
+#### 3. Create Resource Group
+```bash
+az group create --name %RESOURCE_GROUP% --location %LOCATION%
+```
+
+#### 4. Create Container Registry
+```bash
+az acr create --resource-group %RESOURCE_GROUP% --name %ACR_NAME% --sku Basic --admin-enabled true
+```
+
+#### 5. Build Docker Image (Using Azure - No local Docker needed)
+```bash
+az acr build --registry %ACR_NAME% --image sap-finance-dashboard:latest .
+```
+
+#### 6. Create Container Apps Environment
+```bash
+az containerapp env create --name %ENVIRONMENT_NAME% --resource-group %RESOURCE_GROUP% --location %LOCATION%
+```
+
+#### 7. Deploy Container App
+```bash
+az containerapp create ^
+  --name %APP_NAME% ^
+  --resource-group %RESOURCE_GROUP% ^
+  --environment %ENVIRONMENT_NAME% ^
+  --image %ACR_NAME%.azurecr.io/sap-finance-dashboard:latest ^
+  --registry-server %ACR_NAME%.azurecr.io ^
+  --target-port 7862 ^
+  --ingress external ^
+  --cpu 2 ^
+  --memory 4Gi ^
+  --min-replicas 1 ^
+  --max-replicas 3
+```
+
+#### 8. Get Application URL
+```bash
+az containerapp show --name %APP_NAME% --resource-group %RESOURCE_GROUP% --query properties.configuration.ingress.fqdn -o tsv
+```
+
+## Post-Deployment Configuration
+
+### Add Hugging Face Token (Required for AI Features)
+
+1. Go to Azure Portal: https://portal.azure.com
+2. Navigate to your Container App
+3. Go to **Settings** > **Secrets**
+4. Add secret:
+   - Name: `huggingface-token`
+   - Value: Your Hugging Face token (get from https://huggingface.co/settings/tokens)
+5. Go to **Settings** > **Containers** > **Environment variables**
+6. Add environment variable:
+   - Name: `HUGGINGFACE_TOKEN`
+   - Source: Reference a secret
+   - Value: Select `huggingface-token`
+7. Click **Save** and wait for app to restart
+
+### Optional: Add SAP OData Credentials
+
+If connecting to SAP systems:
+- `SAP_ODATA_BASE_URL`: Your SAP OData endpoint
+- `SAP_USERNAME`: SAP username (as secret)
+- `SAP_PASSWORD`: SAP password (as secret)
+
+## Application Architecture
+
+```
+┌─────────────────────────────────────────┐
+│     Azure Container Apps                │
+│  ┌───────────────────────────────────┐  │
+│  │  SAP Finance Dashboard            │  │
+│  │  - Gradio Web Interface           │  │
+│  │  - RPT-1-OSS AI Model             │  │
+│  │  - Port 7862                      │  │
+│  └───────────────────────────────────┘  │
+│              ↕                          │
+│  ┌───────────────────────────────────┐  │
+│  │  Azure Container Registry         │  │
+│  │  - Docker Image Storage           │  │
+│  └───────────────────────────────────┘  │
+└─────────────────────────────────────────┘
+```
+
+## Application Features
+
+Once deployed, you'll have access to:
+
+1. **📊 Dashboard** - Financial metrics and visualizations
+2. **🔍 Data Explorer** - Browse datasets
+3. **📤 Upload** - Upload custom CSV files
+4. **🤖 AI Predictions** - SAP-RPT-1-OSS powered predictions
+5. **🔗 OData** - Connect to SAP systems
+6. **🎮 Playground** - Custom model training
+
+## Scaling Configuration
+
+The default configuration:
+- **Min Replicas:** 1 (always running)
+- **Max Replicas:** 3 (auto-scale under load)
+- **CPU:** 2 cores
+- **Memory:** 4 GB
+
+To adjust scaling:
+```bash
+az containerapp update ^
+  --name %APP_NAME% ^
+  --resource-group %RESOURCE_GROUP% ^
+  --min-replicas 1 ^
+  --max-replicas 5 ^
+  --cpu 4 ^
+  --memory 8Gi
+```
+
+## Cost Optimization
+
+**Estimated Monthly Cost (East US):**
+- Container Apps (1 replica, 2 vCPU, 4GB): ~$60/month
+- Container Registry (Basic): ~$5/month
+- **Total:** ~$65/month
+
+**To reduce costs:**
+1. Set min-replicas to 0 (app sleeps when not in use)
+2. Use smaller CPU/memory allocation
+3. Delete when not needed:
+   ```bash
+   az group delete --name %RESOURCE_GROUP% --yes
+   ```
+
+## Monitoring & Logs
+
+### View Live Logs
+```bash
+az containerapp logs show --name %APP_NAME% --resource-group %RESOURCE_GROUP% --follow
+```
+
+### View Metrics (Azure Portal)
+1. Go to your Container App
+2. Click **Monitoring** > **Metrics**
+3. View:
+   - CPU usage
+   - Memory usage
+   - Request count
+   - Response time
+
+## Updating Your Application
+
+When you make code changes:
+
+```bash
+# Build new image
+az acr build --registry %ACR_NAME% --image sap-finance-dashboard:latest .
+
+# Update container app
+az containerapp update ^
+  --name %APP_NAME% ^
+  --resource-group %RESOURCE_GROUP% ^
+  --image %ACR_NAME%.azurecr.io/sap-finance-dashboard:latest
+```
+
+## Troubleshooting
+
+### App not starting?
+Check logs:
+```bash
+az containerapp logs show --name %APP_NAME% --resource-group %RESOURCE_GROUP% --tail 100
+```
+
+### Out of memory?
+Increase memory allocation:
+```bash
+az containerapp update --name %APP_NAME% --resource-group %RESOURCE_GROUP% --memory 8Gi
+```
+
+### Port issues?
+Verify port is set to 7862:
+```bash
+az containerapp show --name %APP_NAME% --resource-group %RESOURCE_GROUP% --query properties.configuration.ingress
+```
+
+### Can't access URL?
+Check ingress is enabled:
+```bash
+az containerapp ingress show --name %APP_NAME% --resource-group %RESOURCE_GROUP%
+```
+
+## Security Best Practices
+
+1. **Use Managed Identity** for Azure service connections
+2. **Store secrets** in Azure Key Vault
+3. **Enable HTTPS only** (default enabled)
+4. **Restrict ingress** to specific IPs if needed
+5. **Rotate tokens** regularly
+
+## Support
+
+- **Azure Container Apps Docs:** https://learn.microsoft.com/azure/container-apps/
+- **SAP-RPT-1-OSS:** https://github.com/SAP-samples/sap-rpt-1-oss
+- **Gradio Docs:** https://gradio.app/docs
+
+## Clean Up
+
+To delete all resources and stop charges:
+```bash
+az group delete --name %RESOURCE_GROUP% --yes --no-wait
+```
+
+---
+
+**Ready to deploy?** Run the automated script or follow the manual steps above! 🚀

Dockerfile

@@ -0,0 +1,48 @@
+# Dockerfile for SAP Finance Dashboard with RPT-1-OSS Model
+# Optimized for Azure Container Apps deployment
+
+FROM python:3.11-slim
+
+# Set environment variables
+ENV PYTHONDONTWRITEBYTECODE=1
+ENV PYTHONUNBUFFERED=1
+ENV GRADIO_SERVER_NAME=0.0.0.0
+ENV GRADIO_SERVER_PORT=7862
+
+# Install system dependencies
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    git \
+    curl \
+    && rm -rf /var/lib/apt/lists/*
+
+# Set working directory
+WORKDIR /app
+
+# Copy requirements first for better caching
+COPY requirements.txt .
+
+# Install Python dependencies
+RUN pip install --no-cache-dir --upgrade pip && \
+    pip install --no-cache-dir -r requirements.txt
+
+# Install SAP-RPT-1-OSS from GitHub
+RUN pip install --no-cache-dir git+https://github.com/SAP-samples/sap-rpt-1-oss
+
+# Force Gradio 4.x to be installed LAST (override any conflicting dependencies)
+RUN pip install --no-cache-dir --force-reinstall "gradio>=4.0.0"
+
+# Copy application code
+COPY . .
+
+# Create data directory if it doesn't exist
+RUN mkdir -p /app/data
+
+# Expose port
+EXPOSE 7862
+
+# Health check
+HEALTHCHECK --interval=30s --timeout=10s --start-period=120s --retries=3 \
+    CMD curl -f http://localhost:7862/ || exit 1
+
+# Run the application
+CMD ["python", "app_gradio.py"]

README.md

@@ -1,145 +1,37 @@
-# SAP OpenSource Model Playground | SAP-RPT-1-OSS Model Beta
-
-A comprehensive SAP Playground application built with Gradio that integrates the SAP-RPT-1-OSS model for predictive analysis on SAP datasets. Features include synthetic data generation, interactive visualizations, live OData connectivity, AI-powered insights, and a playground for training custom models. 
-<img width="839" height="540" alt="image" src="https://github.com/user-attachments/assets/d6376079-2b83-422a-b4fd-51d517431ec2" />
+---
+title: SAP Finance Dashboard with RPT-1-OSS
+emoji: 📊
+colorFrom: purple
+colorTo: blue
+sdk: gradio
+sdk_version: 4.44.0
+app_file: app_gradio.py
+pinned: false
+license: apache-2.0
+---
+
+# SAP Finance Dashboard with RPT-1-OSS Model
+
+A comprehensive financial dashboard application built with Gradio that integrates the SAP-RPT-1-OSS model for predictive analysis on SAP finance datasets.
 
 ## Features
 
-- **Multiple Synthetic SAP Datasets**: General Ledger accounts, Financial Statements (P&L, Balance Sheet), and Sales Orders
-- **Data Upload**: Upload custom CSV, Parquet, or JSON datasets for analysis
-- **Interactive Visualizations**: Financial charts and graphs using Plotly
-- **SAP-RPT-1-OSS Model Integration**: AI-powered predictions and analysis
-- **Live OData Connection**: Connect to SAP systems to fetch real-time sales order data
-- **Playground Tab**: Upload datasets, configure model parameters, train, and download predictions
-- **Modern UI**: Built with Gradio, a Python-based web framework
-
-## Installation
-
-### Prerequisites
-
-- Python 3.11 or higher
-- Hugging Face account (for SAP model access)
-- SAP OData credentials (optional, for live data connection)
-
-### Setup Steps
-
-1. **Clone the repository**:
-   ```bash
-   git clone <repository-url>
-   cd SAP-RPT-1-OSS-App
-   ```
-
-2. **Create a virtual environment**:
-   ```bash
-   python -m venv venv
-   source venv/bin/activate  # On Windows: venv\Scripts\activate
-   ```
-
-3. **Install dependencies**:
-   ```bash
-   pip install -r requirements.txt
-   ```
-
-4. **Install SAP-RPT-OSS package**:
-   ```bash
-   pip install git+https://github.com/SAP-samples/sap-rpt-1-oss
-   ```
-
-5. **Set up environment variables**:
-   - Copy `.env.example` to `.env`
-   - Fill in your SAP OData credentials
-   - Add your Hugging Face token for model access
-
-6. **Authenticate with Hugging Face**:
-   ```bash
-   pip install huggingface_hub
-   huggingface-cli login
-   ```
-   Or set the `HUGGINGFACE_TOKEN` in your `.env` file.
-
-7. **Generate synthetic data** (optional, if not already generated):
-   ```bash
-   python -c "from utils.data_generator import generate_all_datasets; generate_all_datasets()"
-   ```
+- **Multiple Synthetic SAP Finance Datasets**: General Ledger accounts, Financial Statements, Sales Orders
+- **Interactive Visualizations**: Financial charts using Plotly
+- **SAP-RPT-1-OSS Model Integration**: AI-powered predictions
+- **Playground Tab**: Upload datasets, configure model parameters, train models
 
 ## Usage
 
-### Running the Application
-
-Start the Gradio application:
-
-```bash
-python app_gradio.py
-```
-
-The application will be available at `http://localhost:7862` (default Gradio port).
+1. **Dashboard**: View key financial metrics and visualizations
+2. **Data Explorer**: Browse and filter datasets
+3. **Predictions**: Use SAP-RPT-1-OSS model for AI predictions
+4. **Playground**: Upload custom datasets and train models
 
-### Application Tabs
+## Model
 
-1. **Dashboard**: Overview with key financial metrics and visualizations
-2. **Data Explorer**: Browse and filter datasets (GL, Financial Statements, Sales Orders)
-3. **Upload**: Upload custom CSV datasets for analysis
-4. **Predictions**: Use SAP-RPT-1-OSS model for predictions and analysis with pre-configured scenarios
-5. **OData**: Connect to SAP OData services and fetch live data
-6. **Playground**: Upload datasets, configure model parameters (task type, target column, test split, context size, bagging, GPU), train models, and download predictions
-
-## SAP OData Connection Setup
-
-1. Set the following environment variables in your `.env` file:
-   - `SAP_USERNAME`: Your SAP username
-   - `SAP_PASSWORD`: Your SAP password
-   - `SAP_SERVER`: SAP server URL (default: `https://sapes5.sapdevcenter.com/`)
-   - `SAP_CLIENT`: SAP client number (default: `002`)
-
-2. The OData connector uses the base URL:
-   `https://sapes5.sapdevcenter.com/sap/opu/odata/IWBEP/GWSAMPLE_BASIC`
-
-3. Available endpoints:
-   - Sales Orders: `SalesOrderSet`
-   - Products: `ProductSet`
-   - Line Items: `SalesOrderLineItemSet`
-   - Business Partners: `BusinessPartnerSet`
-
-## Model Configuration
-
-The SAP-RPT-1-OSS model supports both classification and regression tasks. For best performance:
-
-- **Recommended**: GPU with at least 80 GB memory, context size 8192, bagging factor 8
-- **Lightweight**: CPU with context size 2048, bagging factor 1
-
-The application automatically detects available resources and adjusts settings accordingly.
-
-## Project Structure
-
-```
-SAP-RPT-1-OSS-App/
-├── app_gradio.py         # Main Gradio application
-├── models/
-│   └── rpt_model.py      # SAP-RPT-1-OSS model wrapper
-├── data/
-│   ├── synthetic_gl_accounts.csv
-│   ├── synthetic_financial_statements.csv
-│   └── synthetic_sales_orders.csv
-├── utils/
-│   ├── data_generator.py # Generate synthetic SAP finance data
-│   ├── visualizations.py # Chart generation functions
-│   ├── odata_connector.py # OData connection utilities
-│   └── playground.py     # Playground utilities for model training
-├── requirements.txt
-├── README.md
-└── .env.example
-```
+This Space uses the [SAP-RPT-1-OSS](https://huggingface.co/SAP/sap-rpt-1-oss) model for predictions.
 
 ## License
 
-This project is licensed under the Apache Software License, version 2.0.
-
-## Support
-
-For issues or questions, please create an issue in this repository.
-
-## Acknowledgments
-
-- SAP-RPT-1-OSS model: [Hugging Face](https://huggingface.co/SAP/sap-rpt-1-oss)
-- Gradio framework: [Gradio Documentation](https://www.gradio.app/docs/)
-
+Apache 2.0

README_HF.md

@@ -0,0 +1,37 @@
+---
+title: SAP Finance Dashboard with RPT-1-OSS
+emoji: 📊
+colorFrom: purple
+colorTo: blue
+sdk: gradio
+sdk_version: 4.44.0
+app_file: app_gradio.py
+pinned: false
+license: apache-2.0
+---
+
+# SAP Finance Dashboard with RPT-1-OSS Model
+
+A comprehensive financial dashboard application built with Gradio that integrates the SAP-RPT-1-OSS model for predictive analysis on SAP finance datasets.
+
+## Features
+
+- **Multiple Synthetic SAP Finance Datasets**: General Ledger accounts, Financial Statements, Sales Orders
+- **Interactive Visualizations**: Financial charts using Plotly
+- **SAP-RPT-1-OSS Model Integration**: AI-powered predictions
+- **Playground Tab**: Upload datasets, configure model parameters, train models
+
+## Usage
+
+1. **Dashboard**: View key financial metrics and visualizations
+2. **Data Explorer**: Browse and filter datasets
+3. **Predictions**: Use SAP-RPT-1-OSS model for AI predictions
+4. **Playground**: Upload custom datasets and train models
+
+## Model
+
+This Space uses the [SAP-RPT-1-OSS](https://huggingface.co/SAP/sap-rpt-1-oss) model for predictions.
+
+## License
+
+Apache 2.0

app.py

@@ -1,451 +0,0 @@
-"""
-SAP Finance Dashboard with RPT-1-OSS Model
-
-Main Mesop application with multiple pages:
-- Dashboard: Overview with metrics and charts
-- Data Explorer: Browse datasets
-- Upload: Upload custom datasets
-- Predictions: AI-powered predictions using SAP-RPT-1-OSS
-- OData: Connect to SAP OData services
-"""
-
-import os
-import mesop as me
-import pandas as pd
-import numpy as np
-from pathlib import Path
-import json
-import base64
-from typing import Optional, Dict, Any
-import plotly.graph_objects as go
-import plotly.io as pio
-
-# Import utilities
-from utils.data_generator import generate_all_datasets
-from utils.visualizations import (
-    create_revenue_expense_chart,
-    create_balance_sheet_chart,
-    create_gl_summary_chart,
-    create_sales_analytics_chart,
-    create_sales_trend_chart,
-    get_summary_metrics
-)
-from utils.odata_connector import SAPFinanceConnector
-from models.rpt_model import RPTModelWrapper, create_model
-
-# Global state
-from dataclasses import field
-
-@me.stateclass
-class State:
-    gl_data: pd.DataFrame = field(default_factory=pd.DataFrame)
-    financial_data: pd.DataFrame = field(default_factory=pd.DataFrame)
-    sales_data: pd.DataFrame = field(default_factory=pd.DataFrame)
-    uploaded_data: pd.DataFrame = field(default_factory=pd.DataFrame)
-    current_dataset_type: str = ""
-    odata_connector: Optional[SAPFinanceConnector] = None
-    odata_connected: bool = False
-    odata_data: pd.DataFrame = field(default_factory=pd.DataFrame)
-    model_wrapper: Optional[RPTModelWrapper] = None
-    predictions: Optional[np.ndarray] = None
-    prediction_proba: Optional[np.ndarray] = None
-    connection_message: str = ""
-    fetch_message: str = ""
-    model_initialized: bool = False
-    model_trained: bool = False
-
-
-def load_datasets(state: State):
-    """Load synthetic datasets if they exist."""
-    data_dir = Path("data")
-    
-    if (data_dir / "synthetic_gl_accounts.csv").exists():
-        state.gl_data = pd.read_csv(data_dir / "synthetic_gl_accounts.csv")
-    
-    if (data_dir / "synthetic_financial_statements.csv").exists():
-        state.financial_data = pd.read_csv(data_dir / "synthetic_financial_statements.csv")
-    
-    if (data_dir / "synthetic_sales_orders.csv").exists():
-        state.sales_data = pd.read_csv(data_dir / "synthetic_sales_orders.csv")
-
-
-def plotly_to_html(fig_dict: Dict[str, Any]) -> str:
-    """Convert Plotly figure dict to HTML string."""
-    if not fig_dict:
-        return "<p>No chart data available</p>"
-    
-    try:
-        fig = go.Figure(fig_dict)
-        html_str = pio.to_html(fig, include_plotlyjs='cdn', div_id="plotly-div")
-        return html_str
-    except Exception as e:
-        return f"<p>Error rendering chart: {str(e)}</p>"
-
-
-@me.page(path="/", title="SAP Finance Dashboard")
-def dashboard_page():
-    """Main dashboard page with overview metrics and charts."""
-    state = me.state(State)
-    
-    me.text("SAP Finance Dashboard", style=me.Style(font_size=32, font_weight="bold", margin=me.Margin(bottom=16)))
-    
-    # Load datasets if not loaded
-    if state.gl_data.empty and state.financial_data.empty and state.sales_data.empty:
-        load_datasets(state)
-    
-    # Generate datasets if they don't exist
-    if state.gl_data.empty or state.financial_data.empty or state.sales_data.empty:
-        with me.box(style=me.Style(padding=16, background="#fff3cd", border_radius=8, margin=me.Margin(bottom=16))):
-            me.text("Generating synthetic datasets...", style=me.Style(color="#856404"))
-            generate_all_datasets()
-            load_datasets(state)
-    
-    # Summary metrics
-    with me.box(style=me.Style(display="grid", grid_template_columns="repeat(4, 1fr)", gap=16, margin=me.Margin(bottom=24))):
-        if not state.gl_data.empty:
-            gl_metrics = get_summary_metrics(state.gl_data, "gl")
-            with me.box(style=me.Style(padding=16, background="#f8f9fa", border_radius=8)):
-                me.text("GL Transactions", style=me.Style(font_weight="bold"))
-                me.text(f"{gl_metrics.get('Total Transactions', 0):,}")
-        
-        if not state.financial_data.empty:
-            fin_metrics = get_summary_metrics(state.financial_data, "financial")
-            with me.box(style=me.Style(padding=16, background="#f8f9fa", border_radius=8)):
-                me.text("Latest Revenue", style=me.Style(font_weight="bold"))
-                me.text(f"${fin_metrics.get('Latest Revenue', 0):,.2f}")
-        
-        if not state.sales_data.empty:
-            sales_metrics = get_summary_metrics(state.sales_data, "sales")
-            with me.box(style=me.Style(padding=16, background="#f8f9fa", border_radius=8)):
-                me.text("Total Sales", style=me.Style(font_weight="bold"))
-                me.text(f"${sales_metrics.get('Total Sales', 0):,.2f}")
-        
-        with me.box(style=me.Style(padding=16, background="#f8f9fa", border_radius=8)):
-            me.text("Datasets", style=me.Style(font_weight="bold"))
-            count = sum([
-                not state.gl_data.empty,
-                not state.financial_data.empty,
-                not state.sales_data.empty,
-                not state.uploaded_data.empty
-            ])
-            me.text(f"{count} loaded")
-    
-    # Charts
-    if not state.financial_data.empty:
-        with me.box(style=me.Style(margin=me.Margin(bottom=24))):
-            me.text("Financial Trends", style=me.Style(font_size=20, font_weight="bold", margin=me.Margin(bottom=8)))
-            chart_data = create_revenue_expense_chart(state.financial_data)
-            if chart_data:
-                html_chart = plotly_to_html(chart_data)
-                me.html(html_chart)
-    
-    if not state.financial_data.empty:
-        with me.box(style=me.Style(margin=me.Margin(bottom=24))):
-            me.text("Balance Sheet", style=me.Style(font_size=20, font_weight="bold", margin=me.Margin(bottom=8)))
-            chart_data = create_balance_sheet_chart(state.financial_data)
-            if chart_data:
-                html_chart = plotly_to_html(chart_data)
-                me.html(html_chart)
-    
-    if not state.sales_data.empty:
-        with me.box(style=me.Style(margin=me.Margin(bottom=24))):
-            me.text("Sales Analytics", style=me.Style(font_size=20, font_weight="bold", margin=me.Margin(bottom=8)))
-            chart_data = create_sales_analytics_chart(state.sales_data)
-            if chart_data:
-                html_chart = plotly_to_html(chart_data)
-                me.html(html_chart)
-
-
-@me.page(path="/explorer", title="Data Explorer")
-def explorer_page():
-    """Data explorer page to browse and filter datasets."""
-    state = me.state(State)
-    
-    me.text("Data Explorer", style=me.Style(font_size=32, font_weight="bold", margin=me.Margin(bottom=16)))
-    
-    # Dataset selector
-    with me.box(style=me.Style(margin=me.Margin(bottom=16))):
-        me.text("Select Dataset:", style=me.Style(font_weight="bold", margin=me.Margin(bottom=8)))
-        dataset_options = [
-            ("GL Accounts", "gl"),
-            ("Financial Statements", "financial"),
-            ("Sales Orders", "sales"),
-            ("Uploaded Data", "uploaded")
-        ]
-        
-        me.select(
-            label="Dataset",
-            options=[me.SelectOption(label=label, value=value) for label, value in dataset_options],
-            on_selection_change=on_dataset_selection_change,
-            value=state.current_dataset_type
-        )
-    
-    # Display selected dataset
-    if state.current_dataset_type:
-        display_dataset(state, state.current_dataset_type)
-
-
-def on_dataset_selection_change(e: me.SelectSelectionChangeEvent):
-    """Handle dataset selection change."""
-    state = me.state(State)
-    state.current_dataset_type = e.value
-
-
-def display_dataset(state: State, dataset_type: str):
-    """Display the selected dataset."""
-    if dataset_type == "gl" and not state.gl_data.empty:
-        df = state.gl_data
-        me.text(f"GL Accounts ({len(df)} records)", style=me.Style(font_size=20, font_weight="bold", margin=me.Margin(bottom=8)))
-        chart_data = create_gl_summary_chart(df)
-        if chart_data:
-            html_chart = plotly_to_html(chart_data)
-            me.html(html_chart)
-        me.table(data=df.head(100).to_dict("records"), style=me.Style(margin=me.Margin(top=16)))
-    
-    elif dataset_type == "financial" and not state.financial_data.empty:
-        df = state.financial_data
-        me.text(f"Financial Statements ({len(df)} records)", style=me.Style(font_size=20, font_weight="bold", margin=me.Margin(bottom=8)))
-        chart_data = create_revenue_expense_chart(df)
-        if chart_data:
-            html_chart = plotly_to_html(chart_data)
-            me.html(html_chart)
-        me.table(data=df.to_dict("records"), style=me.Style(margin=me.Margin(top=16)))
-    
-    elif dataset_type == "sales" and not state.sales_data.empty:
-        df = state.sales_data
-        me.text(f"Sales Orders ({len(df)} records)", style=me.Style(font_size=20, font_weight="bold", margin=me.Margin(bottom=8)))
-        chart_data = create_sales_trend_chart(df)
-        if chart_data:
-            html_chart = plotly_to_html(chart_data)
-            me.html(html_chart)
-        me.table(data=df.head(100).to_dict("records"), style=me.Style(margin=me.Margin(top=16)))
-    
-    elif dataset_type == "uploaded" and not state.uploaded_data.empty:
-        df = state.uploaded_data
-        me.text(f"Uploaded Data ({len(df)} records)", style=me.Style(font_size=20, font_weight="bold", margin=me.Margin(bottom=8)))
-        me.table(data=df.head(100).to_dict("records"), style=me.Style(margin=me.Margin(top=16)))
-    
-    else:
-        me.text("No data available for this dataset type.", style=me.Style(color="#dc3545"))
-
-
-@me.page(path="/upload", title="Upload Data")
-def upload_page():
-    """Upload page for custom datasets."""
-    state = me.state(State)
-    
-    me.text("Upload Dataset", style=me.Style(font_size=32, font_weight="bold", margin=me.Margin(bottom=16)))
-    
-    with me.box(style=me.Style(margin=me.Margin(bottom=16))):
-        me.text("Upload a CSV file to analyze:", style=me.Style(margin=me.Margin(bottom=8)))
-        me.file_upload(
-            label="Choose CSV File",
-            accept=".csv",
-            on_upload=handle_file_upload
-        )
-    
-    if not state.uploaded_data.empty:
-        me.text("Uploaded Data Preview:", style=me.Style(font_size=20, font_weight="bold", margin=me.Margin(top=16, bottom=8)))
-        me.table(data=state.uploaded_data.head(50).to_dict("records"))
-
-
-def handle_file_upload(e: me.UploadEvent):
-    """Handle file upload."""
-    state = me.state(State)
-    try:
-        if e.files:
-            file = e.files[0]
-            df = pd.read_csv(file.getvalue())
-            state.uploaded_data = df
-    except Exception as ex:
-        pass
-
-
-@me.page(path="/predictions", title="Predictions")
-def predictions_page():
-    """Predictions page using SAP-RPT-1-OSS model."""
-    state = me.state(State)
-    
-    me.text("AI Predictions with SAP-RPT-1-OSS", style=me.Style(font_size=32, font_weight="bold", margin=me.Margin(bottom=16)))
-    
-    with me.box(style=me.Style(margin=me.Margin(bottom=16))):
-        me.text("Model Configuration", style=me.Style(font_size=20, font_weight="bold", margin=me.Margin(bottom=8)))
-        
-        me.select(
-            label="Model Type",
-            options=[
-                me.SelectOption(label="Classifier", value="classifier"),
-                me.SelectOption(label="Regressor", value="regressor")
-            ]
-        )
-        
-        me.checkbox(label="Use GPU (requires 80GB memory)", checked=False)
-        
-        me.button("Initialize Model", on_click=on_init_model)
-    
-    if state.model_initialized:
-        me.text("Model initialized successfully!", style=me.Style(color="#28a745", margin=me.Margin(bottom=16)))
-        
-        # Dataset selection for training
-        with me.box(style=me.Style(margin=me.Margin(bottom=16))):
-            me.text("Select Training Data", style=me.Style(font_weight="bold", margin=me.Margin(bottom=8)))
-            dataset_options = [
-                ("GL Accounts", "gl"),
-                ("Financial Statements", "financial"),
-                ("Sales Orders", "sales"),
-                ("Uploaded Data", "uploaded")
-            ]
-            
-            me.select(
-                label="Dataset",
-                options=[me.SelectOption(label=label, value=value) for label, value in dataset_options]
-            )
-        
-        me.button("Train Model", on_click=on_train_model)
-        
-        if state.model_trained:
-            me.text("Model trained successfully!", style=me.Style(color="#28a745", margin=me.Margin(top=16)))
-        
-        if state.predictions is not None:
-            me.text("Predictions:", style=me.Style(font_size=20, font_weight="bold", margin=me.Margin(top=16, bottom=8)))
-            me.text(str(state.predictions[:10]))  # Show first 10 predictions
-    
-    else:
-        with me.box(style=me.Style(padding=16, background="#fff3cd", border_radius=8)):
-            me.text("Please initialize the model first.", style=me.Style(color="#856404"))
-
-
-def on_init_model(e: me.ClickEvent):
-    """Initialize the model."""
-    state = me.state(State)
-    try:
-        state.model_wrapper = create_model(model_type="classifier", use_gpu=False)
-        state.model_initialized = True
-    except Exception as ex:
-        state.connection_message = f"Error initializing model: {str(ex)}"
-
-
-def on_train_model(e: me.ClickEvent):
-    """Train the model."""
-    state = me.state(State)
-    try:
-        if state.model_wrapper and not state.gl_data.empty:
-            # Simple example: use GL data
-            X = state.gl_data.select_dtypes(include=[np.number]).dropna()
-            if len(X) > 0:
-                # Create a simple target for classification
-                y = (X.iloc[:, 0] > X.iloc[:, 0].median()).astype(int)
-                state.model_wrapper.fit(X, y)
-                state.model_trained = True
-    except Exception as ex:
-        state.connection_message = f"Error training model: {str(ex)}"
-
-
-@me.page(path="/odata", title="OData Connection")
-def odata_page():
-    """OData connection page for SAP data."""
-    state = me.state(State)
-    
-    me.text("SAP OData Connection", style=me.Style(font_size=32, font_weight="bold", margin=me.Margin(bottom=16)))
-    
-    # Connection status
-    with me.box(style=me.Style(margin=me.Margin(bottom=16))):
-        if state.odata_connector is None:
-            state.odata_connector = SAPFinanceConnector()
-        
-        me.button("Test Connection", on_click=on_test_odata_connection)
-        
-        if state.connection_message:
-            color = "#28a745" if state.odata_connected else "#dc3545"
-            me.text(state.connection_message, style=me.Style(color=color, margin=me.Margin(top=8)))
-        elif state.odata_connected:
-            me.text("✓ Connected to SAP OData", style=me.Style(color="#28a745", margin=me.Margin(top=8)))
-        else:
-            me.text("Not connected", style=me.Style(color="#dc3545", margin=me.Margin(top=8)))
-    
-    # Fetch options
-    if state.odata_connected:
-        with me.box(style=me.Style(margin=me.Margin(bottom=16))):
-            me.text("Fetch Data", style=me.Style(font_size=20, font_weight="bold", margin=me.Margin(bottom=8)))
-            
-            top_count = me.number_input(label="Number of records", value=100, min_value=1, max_value=1000)
-            
-            with me.box(style=me.Style(display="grid", grid_template_columns="repeat(2, 1fr)", gap=8, margin=me.Margin(top=8))):
-                me.button("Fetch Sales Orders", on_click=lambda e: on_fetch_odata(e, "orders", top_count))
-                me.button("Fetch Products", on_click=lambda e: on_fetch_odata(e, "products", top_count))
-                me.button("Fetch Line Items", on_click=lambda e: on_fetch_odata(e, "line_items", top_count))
-                me.button("Fetch Partners", on_click=lambda e: on_fetch_odata(e, "partners", top_count))
-        
-        if state.fetch_message:
-            me.text(state.fetch_message, style=me.Style(color="#28a745", margin=me.Margin(top=8)))
-        
-        # Display fetched data
-        if not state.odata_data.empty:
-            me.text("Fetched Data:", style=me.Style(font_size=20, font_weight="bold", margin=me.Margin(top=16, bottom=8)))
-            me.table(data=state.odata_data.head(100).to_dict("records"))
-
-
-def on_test_odata_connection(e: me.ClickEvent):
-    """Test OData connection."""
-    state = me.state(State)
-    try:
-        if state.odata_connector is None:
-            state.odata_connector = SAPFinanceConnector()
-        
-        connected, message = state.odata_connector.test_connection()
-        state.odata_connected = connected
-        state.connection_message = message
-    except Exception as ex:
-        state.connection_message = f"Error: {str(ex)}"
-        state.odata_connected = False
-
-
-def on_fetch_odata(e: me.ClickEvent, entity_type: str, top: int):
-    """Fetch data from OData service."""
-    state = me.state(State)
-    try:
-        if not state.odata_connected:
-            state.fetch_message = "Please connect first!"
-            return
-        
-        if entity_type == "orders":
-            state.odata_data = state.odata_connector.fetch_orders_df(top)
-        elif entity_type == "products":
-            state.odata_data = state.odata_connector.fetch_products_df(top)
-        elif entity_type == "line_items":
-            state.odata_data = state.odata_connector.fetch_line_items_df(top)
-        elif entity_type == "partners":
-            state.odata_data = state.odata_connector.fetch_partners_df(top)
-        
-        state.fetch_message = f"Fetched {len(state.odata_data)} records"
-    except Exception as ex:
-        state.fetch_message = f"Error fetching data: {str(ex)}"
-
-
-# Navigation
-@me.page(path="/nav", title="Navigation")
-def nav_page():
-    """Navigation page."""
-    me.text("Navigation", style=me.Style(font_size=32, font_weight="bold", margin=me.Margin(bottom=16)))
-    
-    nav_links = [
-        ("Dashboard", "/"),
-        ("Data Explorer", "/explorer"),
-        ("Upload", "/upload"),
-        ("Predictions", "/predictions"),
-        ("OData", "/odata")
-    ]
-    
-    for label, path in nav_links:
-        with me.box(style=me.Style(margin=me.Margin(bottom=8))):
-            me.link(label, path=path, style=me.Style(font_size=18, text_decoration="none"))
-
-
-if __name__ == "__main__":
-    # Generate datasets if they don't exist
-    data_dir = Path("data")
-    if not (data_dir / "synthetic_gl_accounts.csv").exists():
-        generate_all_datasets()
-    
-    me.run()
-

app_gradio.py

@@ -1415,8 +1415,14 @@ with gr.Blocks(title="SAP Finance Dashboard", theme=gr.themes.Soft(), css="""
 
 
 if __name__ == "__main__":
+    import os
+    
     # Load datasets on startup
     load_datasets()
     
+    # Get server configuration from environment variables (for container deployment)
+    server_name = os.environ.get("GRADIO_SERVER_NAME", "0.0.0.0")
+    server_port = int(os.environ.get("GRADIO_SERVER_PORT", 7862))
+    
     # Launch the app
-    app.launch(share=False, server_name="127.0.0.1", server_port=7862, quiet=False)
+    app.launch(share=False, server_name=server_name, server_port=server_port, quiet=False)

deploy-azure.ps1

@@ -0,0 +1,135 @@
+# Azure Container Apps Deployment Script for SAP Finance Dashboard
+# Run this script in PowerShell after logging in to Azure CLI
+
+# ============================================
+# CONFIGURATION - Update these values
+# ============================================
+$RESOURCE_GROUP = "rg-sap-finance-dashboard"
+$LOCATION = "eastus"
+$ACR_NAME = "acrsapfinance$(Get-Random -Maximum 9999)"  # Must be globally unique
+$ENVIRONMENT_NAME = "env-sap-finance"
+$APP_NAME = "sap-finance-dashboard"
+
+# Optional: Set your secrets (or configure later in Azure Portal)
+$HUGGINGFACE_TOKEN = ""  # Your Hugging Face token
+$SAP_USERNAME = ""       # SAP OData username
+$SAP_PASSWORD = ""       # SAP OData password
+
+# ============================================
+# STEP 1: Login to Azure (if not already logged in)
+# ============================================
+Write-Host "Step 1: Checking Azure login..." -ForegroundColor Cyan
+$account = az account show 2>$null | ConvertFrom-Json
+if (-not $account) {
+    Write-Host "Please login to Azure..." -ForegroundColor Yellow
+    az login
+}
+Write-Host "Logged in as: $($account.user.name)" -ForegroundColor Green
+
+# ============================================
+# STEP 2: Create Resource Group
+# ============================================
+Write-Host "`nStep 2: Creating Resource Group..." -ForegroundColor Cyan
+az group create --name $RESOURCE_GROUP --location $LOCATION
+Write-Host "Resource Group '$RESOURCE_GROUP' created in $LOCATION" -ForegroundColor Green
+
+# ============================================
+# STEP 3: Create Azure Container Registry
+# ============================================
+Write-Host "`nStep 3: Creating Azure Container Registry..." -ForegroundColor Cyan
+az acr create --resource-group $RESOURCE_GROUP --name $ACR_NAME --sku Basic --admin-enabled true
+Write-Host "Container Registry '$ACR_NAME' created" -ForegroundColor Green
+
+# Get ACR credentials
+$ACR_USERNAME = az acr credential show --name $ACR_NAME --query username -o tsv
+$ACR_PASSWORD = az acr credential show --name $ACR_NAME --query "passwords[0].value" -o tsv
+$ACR_LOGIN_SERVER = az acr show --name $ACR_NAME --query loginServer -o tsv
+
+Write-Host "ACR Login Server: $ACR_LOGIN_SERVER" -ForegroundColor Yellow
+
+# ============================================
+# STEP 4: Build and Push Docker Image
+# ============================================
+Write-Host "`nStep 4: Building and pushing Docker image..." -ForegroundColor Cyan
+Write-Host "This may take several minutes..." -ForegroundColor Yellow
+
+# Build using ACR Tasks (no local Docker required)
+az acr build --registry $ACR_NAME --image "${APP_NAME}:latest" .
+
+Write-Host "Docker image built and pushed to ACR" -ForegroundColor Green
+
+# ============================================
+# STEP 5: Create Container Apps Environment
+# ============================================
+Write-Host "`nStep 5: Creating Container Apps Environment..." -ForegroundColor Cyan
+az containerapp env create `
+    --name $ENVIRONMENT_NAME `
+    --resource-group $RESOURCE_GROUP `
+    --location $LOCATION
+
+Write-Host "Container Apps Environment '$ENVIRONMENT_NAME' created" -ForegroundColor Green
+
+# ============================================
+# STEP 6: Deploy Container App
+# ============================================
+Write-Host "`nStep 6: Deploying Container App..." -ForegroundColor Cyan
+
+# Build the secrets parameter if tokens are provided
+$secretsParam = ""
+$envVarsParam = ""
+
+if ($HUGGINGFACE_TOKEN) {
+    $secretsParam = "--secrets huggingface-token=$HUGGINGFACE_TOKEN"
+    $envVarsParam = "--env-vars HUGGINGFACE_TOKEN=secretref:huggingface-token"
+}
+
+# Create the container app
+az containerapp create `
+    --name $APP_NAME `
+    --resource-group $RESOURCE_GROUP `
+    --environment $ENVIRONMENT_NAME `
+    --image "${ACR_LOGIN_SERVER}/${APP_NAME}:latest" `
+    --registry-server $ACR_LOGIN_SERVER `
+    --registry-username $ACR_USERNAME `
+    --registry-password $ACR_PASSWORD `
+    --target-port 7862 `
+    --ingress external `
+    --cpu 2 `
+    --memory 4Gi `
+    --min-replicas 1 `
+    --max-replicas 3 `
+    --query properties.configuration.ingress.fqdn
+
+Write-Host "`nContainer App deployed!" -ForegroundColor Green
+
+# ============================================
+# STEP 7: Get Application URL
+# ============================================
+Write-Host "`nStep 7: Getting Application URL..." -ForegroundColor Cyan
+$APP_URL = az containerapp show `
+    --name $APP_NAME `
+    --resource-group $RESOURCE_GROUP `
+    --query properties.configuration.ingress.fqdn -o tsv
+
+Write-Host "`n============================================" -ForegroundColor Green
+Write-Host "DEPLOYMENT COMPLETE!" -ForegroundColor Green
+Write-Host "============================================" -ForegroundColor Green
+Write-Host "`nYour SAP Finance Dashboard is available at:" -ForegroundColor Cyan
+Write-Host "https://$APP_URL" -ForegroundColor Yellow
+Write-Host "`n============================================" -ForegroundColor Green
+
+# ============================================
+# NEXT STEPS
+# ============================================
+Write-Host "`nNEXT STEPS:" -ForegroundColor Cyan
+Write-Host "1. Configure secrets in Azure Portal:" -ForegroundColor White
+Write-Host "   - HUGGINGFACE_TOKEN: Your Hugging Face API token" -ForegroundColor Gray
+Write-Host "   - SAP_USERNAME: SAP OData username (optional)" -ForegroundColor Gray
+Write-Host "   - SAP_PASSWORD: SAP OData password (optional)" -ForegroundColor Gray
+Write-Host "`n2. To update the app after code changes:" -ForegroundColor White
+Write-Host "   az acr build --registry $ACR_NAME --image ${APP_NAME}:latest ." -ForegroundColor Gray
+Write-Host "   az containerapp update --name $APP_NAME --resource-group $RESOURCE_GROUP --image ${ACR_LOGIN_SERVER}/${APP_NAME}:latest" -ForegroundColor Gray
+Write-Host "`n3. To view logs:" -ForegroundColor White
+Write-Host "   az containerapp logs show --name $APP_NAME --resource-group $RESOURCE_GROUP --follow" -ForegroundColor Gray
+Write-Host "`n4. To delete all resources:" -ForegroundColor White
+Write-Host "   az group delete --name $RESOURCE_GROUP --yes --no-wait" -ForegroundColor Gray

requirements.txt

@@ -1,4 +1,3 @@
-mesop>=0.4.0
 pandas>=2.0.0
 numpy>=1.24.0
 #sap-rpt-oss
@@ -7,7 +6,7 @@ plotly>=5.17.0
 requests>=2.31.0
 python-dotenv>=1.0.0
 scikit-learn>=1.3.0
-gradio
+gradio>=4.0.0
 pyarrow>=10.0.0
 pyzmq>=25.0.0