Merge pull request #58 from TuanaCelik/main

LlamaCloud example

Author: cristipufu

samples/quickstart-llamacloud/.env.example

@@ -0,0 +1,16 @@
+# OpenAI Configuration
+OPENAI_API_KEY=your_openai_api_key_here
+
+# LlamaCloud Configuration
+LLAMACLOUD_API_KEY=your_llamacloud_api_key
+LLAMACLOUD_ORG_ID=your_llamacloud_org_id
+LLAMACLOUD_PROJECT_NAME=your_llamacloud_project_name
+
+# LlamaCloud Index Configuration
+# Index 1: Company Travel Policies and Rates
+LLAMACLOUD_INDEX_1_NAME=company-policies
+LLAMACLOUD_INDEX_1_DESCRIPTION="Information about company travel rates per states/cities and general company policy"
+
+# Index 2: Personal Travel Preferences
+LLAMACLOUD_INDEX_2_NAME=personal-preferences
+LLAMACLOUD_INDEX_2_DESCRIPTION="Information about user's personal preferences"

samples/quickstart-llamacloud/README.md

@@ -0,0 +1,168 @@
+# Quickstart LlamaCloud Agent
+
+This project demonstrates how to integrate UiPath with LlamaIndex and LlamaCloud for document search and travel assistance workflows using a FunctionAgent.
+
+## Overview
+
+The Quickstart LlamaCloud Agent provides a FunctionAgent that can:
+1. Search company travel policies and rates from the `company-policy` index
+2. Search user's personal travel preferences from the `personal-preferences` index
+3. Generate comprehensive travel recommendations combining both sources
+4. Deploy as a UiPath agent for automation workflows
+
+## Features
+
+- **FunctionAgent Architecture**: Uses LlamaIndex's FunctionAgent for intelligent tool selection
+- **Dual Index Search**: Searches both company policies and personal preferences
+- **Smart Tool Selection**: Automatically chooses the right tools based on user queries
+- **Comprehensive Travel Guidance**: Combines policy and preference information
+- **UiPath Integration**: Ready for deployment to UiPath Cloud
+
+## Prerequisites
+
+- Python 3.10+
+- UiPath Cloud account
+- OpenAI API key
+- LlamaCloud account with API access
+
+## Setup
+
+
+### 1. Set Up Virtual Environment
+
+We recommend using `uv` for package management:
+
+```bash
+# Initialize a new uv project
+uv init .
+
+# Create and activate virtual environment
+uv venv
+source .venv/bin/activate  # On Windows: .venv\Scripts\activate
+```
+
+Alternatively, use `pip`:
+
+```bash
+# Create virtual environment
+python -m venv .venv
+
+# Activate virtual environment
+# On Windows PowerShell: .venv\Scripts\Activate.ps1
+# On Windows CMD: .venv\Scripts\activate
+# On macOS/Linux: source .venv/bin/activate
+source .venv/bin/activate
+
+# Upgrade pip
+python -m pip install --upgrade pip
+```
+
+### 2. Install Dependencies
+
+```bash
+# Install the project dependencies
+pip install -e .
+```
+
+### 3. Configure Environment Variables
+
+```bash
+# Copy the example environment file
+cp .env.example .env
+
+# Edit .env with your actual API keys and configuration
+```
+
+Required environment variables:
+- `OPENAI_API_KEY`: Your OpenAI API key
+- `LLAMACLOUD_API_KEY`: Your LlamaCloud API key
+- `LLAMACLOUD_ORG_ID`: Your LlamaCloud organization ID
+- `LLAMACLOUD_PROJECT_NAME`: Your LlamaCloud project name
+- `LLAMACLOUD_INDEX_1_NAME`: First index name (e.g., "company_policy")
+- `LLAMACLOUD_INDEX_2_NAME`: Second index name (e.g., "personal_preferences")
+
+### 4. Configure LlamaCloud Indexes
+
+- Set up your indexes in LlamaCloud with the names specified in your `.env` file
+- You can drag and drop the files in `/sample_data` into the relevant indexes
+- Update the project name and organization ID in your `.env` file
+- Ensure your API key has access to the specified indexes
+
+## Available Functions
+
+The agent has access to three main functions:
+
+### 1. `search_company_policy(query: str)`
+Searches the company policy index for travel rates, guidelines, and company policies.
+- **Use case**: "What are the travel rates for New York?"
+- **Returns**: Company policy information with source files and relevance scores
+
+### 2. `search_personal_preferences(query: str)`
+Searches the personal preferences index for user's travel preferences and requirements.
+- **Use case**: "What are my travel preferences?"
+- **Returns**: Personal preference information with source files and relevance scores
+
+### 3. `get_travel_recommendation(query: str)`
+Generates comprehensive travel recommendations combining both company policies and personal preferences.
+- **Use case**: "Give me travel recommendations for a business trip"
+- **Returns**: Combined analysis from both indexes
+
+## Usage
+
+### Local Testing
+
+
+#### Using UiPath CLI (Recommended)
+
+```bash
+# Run the agent with a query
+uipath run agent '{"query": "What are the travel rates for New York?"}'
+```
+
+#### Option 3: Using Input File
+
+Create an `input.json` file:
+
+```json
+{
+  "user_msg": "What are the travel rates for California?"
+}
+```
+
+
+### UiPath Deployment
+
+1. **Authenticate with UiPath:**
+   ```bash
+   uipath auth
+   ```
+
+2. **Package your project:**
+   ```bash
+   uipath pack
+   ```
+
+3. **Publish to UiPath Cloud:**
+   ```bash
+   uipath publish --my-workspace
+   ```
+
+4. **Invoke the agent:**
+   ```bash
+   uipath invoke agent '{"user_msg": "What are the travel rates for New York?"}'
+   ```
+
+
+## Example Queries
+
+- "What are the travel rates for California?"
+- "What are my travel preferences?"
+- "Give me travel recommendations for a business trip to New York"
+- "What are the company travel guidelines?"
+- "What are my personal travel requirements?"
+
+## Dependencies
+
+- `uipath-llamaindex`: UiPath LlamaIndex integration
+- `llama-index-llms-openai`: OpenAI LLM integration
+- `llama-cloud-services`: LlamaCloud integration 

samples/quickstart-llamacloud/llama_index.json

@@ -0,0 +1,7 @@
+{
+  "dependencies": ["."],
+  "workflows": {
+    "agent": "main.py:agent"
+  },
+  "env": ".env"
+} 

samples/quickstart-llamacloud/main.py

@@ -0,0 +1,145 @@
+import os
+from typing import List
+from dotenv import load_dotenv
+
+from llama_index.core.agent.workflow import FunctionAgent
+from llama_index.llms.openai import OpenAI
+from llama_index.indices.managed.llama_cloud import LlamaCloudIndex
+
+load_dotenv()
+
+# Initialize LlamaCloud Index connections
+company_policy_index = LlamaCloudIndex(
+    name=os.getenv("LLAMACLOUD_INDEX_1_NAME"),
+    project_name=os.getenv("LLAMACLOUD_PROJECT_NAME"),
+    organization_id=os.getenv("LLAMACLOUD_ORG_ID"),
+    api_key=os.getenv("LLAMACLOUD_API_KEY")
+)
+
+personal_preferences_index = LlamaCloudIndex(
+    name=os.getenv("LLAMACLOUD_INDEX_2_NAME"),
+    project_name=os.getenv("LLAMACLOUD_PROJECT_NAME"),
+    organization_id=os.getenv("LLAMACLOUD_ORG_ID"),
+    api_key=os.getenv("LLAMACLOUD_API_KEY")
+)
+
+llm = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))
+
+
+def search_company_policy(query: str) -> str:
+    """
+    Search the company policy index for travel rates, guidelines, and company policies.
+    
+    Args:
+        query (str): The search query about company travel policies, rates, or guidelines
+        
+    Returns:
+        str: Relevant information from company policy documents
+    """
+    try:
+        retriever = company_policy_index.as_retriever()
+        retrieved_nodes = retriever.retrieve(query)
+        
+        if not retrieved_nodes:
+            return "No relevant company policy information found for your query."
+        
+        # Format the search results
+        results = []
+        for i, node in enumerate(retrieved_nodes[:3], 1):  # Limit to top 3 results
+            results.append(f"Result {i}:")
+            results.append(f"  Content: {node.text[:300]}...")
+            results.append(f"  Relevance Score: {node.score:.4f}")
+            if hasattr(node.node, 'metadata'):
+                metadata = node.node.metadata
+                if 'file_name' in metadata:
+                    results.append(f"  Source: {metadata['file_name']}")
+                if 'page_label' in metadata:
+                    results.append(f"  Page: {metadata['page_label']}")
+            results.append("")
+        
+        return "\n".join(results)
+    except Exception as e:
+        return f"Error searching company policy index: {str(e)}"
+
+
+def search_personal_preferences(query: str) -> str:
+    """
+    Search the personal preferences index for user's travel preferences and requirements.
+    
+    Args:
+        query (str): The search query about personal travel preferences or requirements
+        
+    Returns:
+        str: Relevant information from personal preference documents
+    """
+    try:
+        retriever = personal_preferences_index.as_retriever()
+        retrieved_nodes = retriever.retrieve(query)
+        
+        if not retrieved_nodes:
+            return "No relevant personal preference information found for your query."
+        
+        # Format the search results
+        results = []
+        for i, node in enumerate(retrieved_nodes[:3], 1):  # Limit to top 3 results
+            results.append(f"Result {i}:")
+            results.append(f"  Content: {node.text[:300]}...")
+            results.append(f"  Relevance Score: {node.score:.4f}")
+            if hasattr(node.node, 'metadata'):
+                metadata = node.node.metadata
+                if 'file_name' in metadata:
+                    results.append(f"  Source: {metadata['file_name']}")
+                if 'page_label' in metadata:
+                    results.append(f"  Page: {metadata['page_label']}")
+            results.append("")
+        
+        return "\n".join(results)
+    except Exception as e:
+        return f"Error searching personal preferences index: {str(e)}"
+
+
+def get_travel_recommendation(query: str) -> str:
+    """
+    Get travel recommendations based on company policies and personal preferences.
+    
+    Args:
+        query (str): The travel-related query or request for recommendations
+        
+    Returns:
+        str: Travel recommendations combining policy and preference information
+    """
+    try:
+        # Search both indexes for comprehensive information
+        policy_info = search_company_policy(query)
+        preference_info = search_personal_preferences(query)
+        
+        recommendation = f"""
+Travel Recommendation Analysis:
+
+Company Policy Information:
+{policy_info}
+
+Personal Preference Information:
+{preference_info}
+
+Based on the above information, here are the key considerations for your travel request.
+        """
+        
+        return recommendation.strip()
+    except Exception as e:
+        return f"Error generating travel recommendation: {str(e)}"
+
+
+# Create the FunctionAgent with our tools
+agent = FunctionAgent(
+    tools=[search_company_policy, search_personal_preferences, get_travel_recommendation],
+    llm=llm,
+    system_prompt="""You are a helpful travel assistant that can search through company travel policies and personal preferences to provide comprehensive travel guidance. 
+
+You have access to three main functions:
+1. search_company_policy - Search for company travel rates, guidelines, and policies
+2. search_personal_preferences - Search for user's personal travel preferences and requirements  
+3. get_travel_recommendation - Get comprehensive travel recommendations combining both sources
+
+Use these tools to help users with travel-related queries, ensuring you provide accurate information from both company policies and personal preferences when relevant."""
+) 

samples/quickstart-llamacloud/pyproject.toml

@@ -0,0 +1,12 @@
+[project]
+name = "quickstart-llamacloud"
+version = "0.0.2"
+description = "UiPath LlamaIndex Quickstart with LlamaCloud Integration"
+authors = [{ name = "Tuana Celik", email = "tuana@runllama.ai" }]
+readme = { file = "README.md", content-type = "text/markdown" }
+requires-python = ">=3.10"
+dependencies = [
+    "uipath-llamaindex>=0.0.11",
+    "llama-index-llms-openai>=0.2.2",
+    "llama-cloud-services"
+]