🏕️ Camping La Rosaleda - AI Receptionist Agent

An intelligent virtual receptionist for Camping La Rosaleda, built with Google ADK (Agent Development Kit) and Vertex AI RAG Engine.

🛠️ Tech Stack

Category	Technology
AI Framework
LLM
RAG Engine
Deployment
Observability
Frontend
Package Manager

📋 Features

• RAG-Powered Responses: Uses Vertex AI RAG Engine to retrieve accurate camping information
• Exact Pricing: Provides precise prices based on season (Low/Medium/High)
• Multi-Channel Deployment:
  • 🌐 Streamlit Web UI (Cloud Run)
  • 🤖 Vertex AI Agent Engine API
  • 💻 Local development with adk web
• Observability: Full tracing with Arize AI (EU region supported)
• Spanish Language: Fully localized for Spanish-speaking users

🏗️ Architecture

┌─────────────────┐     ┌──────────────────┐     ┌─────────────────┐
│   Streamlit UI  │────▶│   Google ADK     │────▶│  Vertex AI RAG  │
│   (Cloud Run)   │     │   Agent Runner   │     │     Engine      │
└─────────────────┘     └──────────────────┘     └─────────────────┘
                               │                         │
                               ▼                         ▼
                        ┌──────────────┐         ┌─────────────────┐
                        │  Gemini 2.0  │         │  RAG Corpus     │
                        │    Flash     │         │  (Camping Data) │
                        └──────────────┘         └─────────────────┘
                               │
                               ▼
                        ┌──────────────┐
                        │   Arize AI   │
                        │   Tracing    │
                        └──────────────┘

🚀 Quick Start

Prerequisites

• Python 3.11+
• UV package manager
• Google Cloud account with billing enabled
• gcloud CLI installed and configured

Installation

1. Clone the repository

   git clone https://github.com/Acquarts/google-adk-rag-agent-camping-receptionist.git
   cd google-adk-rag-agent-camping-receptionist

2. Install dependencies

   uv sync

3. Configure environment variables

   cp .env-example .env
   # Edit .env with your credentials

4. Run locally

   # Option 1: ADK Web Interface
   uv run adk web
   
   # Option 2: Streamlit UI
   uv run streamlit run app.py

⚙️ Configuration

Create a .env file with the following variables:

# Google Cloud
GOOGLE_GENAI_USE_VERTEXAI=1
GOOGLE_CLOUD_PROJECT=your-project-id
GOOGLE_CLOUD_LOCATION=europe-west1

# RAG Corpus
RAG_CORPUS=projects/your-project/locations/europe-west1/ragCorpora/your-corpus-id

# Staging bucket for deployment
STAGING_BUCKET=gs://your-bucket-name

# (Optional) Arize Tracing
ARIZE_SPACE_ID=your-space-id
ARIZE_API_KEY=your-api-key

📦 Deployment

Deploy to Cloud Run (Streamlit UI)

gcloud run deploy camping-la-rosaleda \
  --source . \
  --region europe-west1 \
  --allow-unauthenticated \
  --set-env-vars "GOOGLE_GENAI_USE_VERTEXAI=1,GOOGLE_CLOUD_PROJECT=your-project,GOOGLE_CLOUD_LOCATION=europe-west1,RAG_CORPUS=your-corpus"

Deploy to Vertex AI Agent Engine

uv run python deployment/deploy.py

🗓️ Camping Seasons & Pricing

Season	Dates
Low	Jan 1 - Apr 30, Oct 1 - Dec 31
Medium	May 1 - Jul 10, Aug 16 - Sep 30
High	Jul 11 - Aug 15

The agent automatically calculates prices based on the current season and provides exact totals for reservations.

📁 Project Structure

├── app.py                 # Streamlit UI
├── rag/
│   ├── agent.py           # Main agent definition
│   ├── prompts.py         # System prompts with pricing
│   └── tracing.py         # Arize observability setup
├── deployment/
│   └── deploy.py          # Vertex AI Agent Engine deployment
├── Dockerfile             # Cloud Run container
└── pyproject.toml         # Dependencies

🔍 Observability

This project uses Arize AI for LLM tracing and observability. Traces are sent to the EU endpoint (otlp.eu-west-1a.arize.com) for GDPR compliance.

📄 License

Apache License 2.0

👤 Author

Adrian Acquaroni

• GitHub: @Acquarts

---

Made with ❤️ using Google ADK and Vertex AI