Tunisia-AI-Uni-Tour-2026-Agentic-AI-Workflows-on-Google-Vertex-AI-Workshops-Series

This repository provides hands-on workshop materials for building agentic AI workflows using Google Vertex AI. It covers end-to-end concepts from local agents to cloud-scale orchestration, with a focus on explainability (xAI) and responsible AI. You’ll find demonstrations of a local agent running with Ollama and Python, integration patterns with Vertex AI Agent Builder, and guidance on deploying agentic workflows via Vertex AI Pipelines, models, and feature stores. Use the material to run step-by-step experiments, reproduce the demonstrations, and adapt the architectures to your own agent-driven projects in production environments.

Prerequisites:

• Access to Google Cloud with Vertex AI enabled and appropriate permissions
• Python 3.8+ and pip
• Optional: Ollama installed for local LLM inference (Demo 1)
• git to clone the repository

Installation steps:

1. Clone the repository:

git clone https://github.com/your-org/ Tunisia-AI-Uni-Tour-2026-Agentic-AI-Workflows-on-Google-Vertex-AI-Workshops-Series.git
cd Tunisia-AI-Uni-Tour-2026-Agentic-AI-Workflows-on-Google-Vertex-AI-Workshops-Series

2. Install Python dependencies (if demos use Python):

pip install -r requirements.txt

3. Set up Google Cloud SDK and authenticate:

gcloud auth login
gcloud config set project YOUR_PROJECT_ID

4. Follow the Vertex AI setup guides in the repository to configure Vertex AI Agent Builder, Vertex AI Models, and Pipelines. This typically involves enabling APIs and creating a GCS bucket for artifacts.
5. For local experiments (Demo 1), ensure Ollama is installed and running if you plan to execute the local agent workflow as described in the materials.

Notes:

• The repository contains multiple demos and connectors; refer to the README sections for specific demo prerequisites and runtime steps.

Tips and common issues:

• Ensure you have the correct Google Cloud permissions for Vertex AI resources (Agent Builder, Models, Pipelines, Feature Store).
• When using Vertex AI pipelines, verify your GCS bucket permissions and region compatibility.
• For explainability (xAI) components, ensure logging and auditability are enabled to meet governance requirements.
• If you encounter authentication errors, re-authenticate with gcloud and verify that the service account used for Vertex AI has the necessary roles.
• The architecture emphasizes separation of planning, memory, and action; when adapting, keep this separation to maintain explainability and testability.