Building an Autonomous Multi-Tool AI Agent on Google Cloud with Vertex AI

A developer detailed the architecture and implementation of an autonomous multi-tool AI agent built on Google Cloud's Vertex AI, using Python and Cloud Run. The agent uses function calling to query inventory data and returns grounded responses, with deployment via Google Cloud Run.

Generative AI applications are rapidly moving beyond single-turn conversational chatbots toward Autonomous Multi-Tool AI Agents . Instead of just generating static text, modern agents evaluate user prompts, make routing decisions, select specialised external tools, and fetch dynamic real-time data before returning a grounded response. In this article, we will break down the end-to-end architecture and implementation of an autonomous agent built using Vertex AI , Python, and Google Cloud infrastructure. High-Level System Architecture The solution uses a three-tier agentic architecture designed for low latency, modularity, and strict session isolation: To start, configure your Google Cloud project and enable the necessary service APIs in Cloud Shell: bash export PROJECT ID=$ gcloud config get-value project export REGION="us-central1" gcloud services enable \ aiplatform.googleapis.com \ run.googleapis.com \ cloudbuild.googleapis.com \ firestore.googleapis.com 1. Defining Agent Tools and Schema Declarations import vertexai from vertexai.generative models import GenerativeModel, FunctionDeclaration, Tool vertexai.init project="YOUR PROJECT ID", location="us-central1" inventory func = FunctionDeclaration name="query inventory", description="Look up product stock, availability, and unit pricing dynamically.", parameters={ "type": "object", "properties": { "item name": { "type": "string", "description": "The specific item or product name to search" }, "category": { "type": "string", "description": "Item category, e.g., beverages, snacks, merchandise" } }, "required": "item name" }, agent tools = Tool function declarations= inventory func plaintext 2. Implementing the Orchestration Logic def query inventory item name: str, category: str = None - dict: Simulated database lookup or Firestore Vector retrieval return { "item": item name, "in stock": True, "quantity": 42, "price usd": 4.50 } model = GenerativeModel model name="gemini-1.5-flash-001", tools= agent tools chat = model.start chat response = chat.send message "Do we have any Cold Brew in stock?" for part in response.candidates 0 .content.parts: if part.function call: fn name = part.function call.name fn args = dict part.function call.args if fn name == "query inventory": tool result = query inventory fn args Return tool output back to the model for final synthesis final response = chat.send message vertexai.generative models.Part.from function response name=fn name, response={"content": tool result} print final response.text plaintext 3. Packaging and Deploying to Google Cloud Run FROM python:3.11-slim WORKDIR /app COPY requirements.txt . RUN pip install --no-cache-dir -r requirements.txt COPY . . EXPOSE 8080 CMD "streamlit", "run", "app.py", "--server.port=8080", "--server.address=0.0.0.0" shell Deploy directly using the Google Cloud CLI: gcloud run deploy genai-agent-service \ --source . \ --region us-central1 \ --allow-unauthenticated