How to use this repository

This repository contains exercises to help you deploy a RAG application in GCP.

Expectations from Dauphine AI Students

Evaluation Criteria

• Code Quality: Add Flake8 and Pylint linters to your settings.json and use the warnings to improve your code.
• Clean Code Principles:
  • Follow the PEP8 standard.
  • Adopt the following practices:
    • Use type hints.
    • Use docstrings only if needed.
    • Utilize classes and functions.
    • Ensure each function performs a single task.
• GCP Deployed Applications:
  • The application should be deployed in GCP.
  • The interface should function as expected.
  • Avoid deploying multiple applications simultaneously.

How to Submit Your Work

• Create a branch with your name : git checkout -b <your-name>.
• Commit your work in the branch : git add . && git commit -m "Your message".
• Push your branch to the repository : git push origin <your-name>.
• Create a pull request from your branch to the dev branch.

Mandatory Steps

• Prefix anything you deploy or create in GCP with your initials.

Prerequisites

• Python 3.11 or higher
• Conda or Miniconda
• Docker
• VS Code
• GCloud SDK
• Cloud SQL Proxy

Create Your Virtual Environment

conda create --name genai_gcp python=3.11
conda activate genai_gcp
pip install -r requirements.txt

Authenticate on GCP

gcloud init # Use your email and project settings
gcloud auth login
gcloud auth application-default login # Authenticate on GCP with admin account

Commands Used ONLY by the Teacher to Create Required Objects in GCP

Configure API, IAM, and Service Account

gcloud init # Use your email and project settings
gcloud auth application-default login # Authenticate on GCP with admin account
gcloud storage buckets create gs://<my-tfstate-bucket> --project=<my-project-id> --location=<my-region> # You need storage.buckets.create permission (ex: roles/editor)
cd terraform/bootstrap
vim backend.tf
# Edit <my-tfstate-bucket> with the name of the bucket created above
vim terraform.tfvars
# Edit the values of project_id, location, artifactregistry_name, and service_account_name
terraform init # You need storage.buckets.list permission (ex: roles/storage.objectUser)
terraform apply -var gcp_account_email="<my-user-email>"
# Return the service account to be copied next line
export GOOGLE_IMPERSONATE_SERVICE_ACCOUNT=<my-service-account>@<my-project-id>.iam.gserviceaccount.com

About the Exercises

This repository contains exercises to help you deploy a RAG application in GCP. Complete the exercises in the following order:

TP Number	Link	Description
1	Open exercice	Create a basic Streamlit app in Cloud Run.
2	Open exercice	Split the backend and frontend of your app and deploy in Cloud Run.
3	Open exercice	Use Gemini LLM to create a basic question-answering app without external knowledge.
4	Open exercice	Create and fill a Cloud SQL instance with the data needed for the RAG model: Load the data in Google Cloud Storage. Chunk the data with Google Cloud Functions and add it to the Cloud SQL instance. Update the data with Google Cloud Functions when new data is added to the bucket.
5	Open exercice	Use this data hosted in Cloud SQL to improve your question-answering app.
6	Open exercice	Use the RAG model to improve your question-answering app.
7	Open exercice	Collect user feedback from your app and store it in BigQuery.
8	Open exercice	Explore alternative architectures for deploying the RAG model in production, such as using Dialogflow.

Roadmap

• Terraform Cloud SQL instance and DB ingestion.
• Terraform User IAM.

Troubleshooting

You will encounter many errors during development. Don't worry: it's part of the learning process. Here are a few hints on how to solve them:

- ERROR: Cannot connect to the Docker daemon # Docker app is not opened
- An env variable is not recognized # Check the .env file, run source .env in the terminal, load_dotenv() in the Python file

Role errors:

# PERMISSION_DENIED: Permission 'iam.serviceAccounts.getIamPolicy' denied on resource
# Command can only be done by an admin
gcloud iam service-accounts add-iam-policy-binding 1021317796643-compute@developer.gserviceaccount.com \
    --member="user:bastinflorian1@gmail.com" \
    --role="roles/iam.serviceAccountUser"