ng-implementation/README.md

# RAG Pattern :: Backend

This repository hosts the code for the frontend application built with Streamlit. The deployment process involves building a Docker image, pushing it to Amazon Elastic Container Registry (ECR), and running it as an ECS task using Terraform for infrastructure management.

## Deployment Workflow

- :one: *Docker Image Build and Push to ECR*: The frontend application is containerized using Docker. The GitLab CI/CD pipeline builds the image and pushes it to an Amazon ECR repository.
- :two: *Infrastructure Provisioning with Terraform*: The necessary AWS infrastructure, including the ECR repository, ECS cluster, and networking resources, is defined and managed using Terraform.
- :three: *ECS Task Deployment*: The Docker image from ECR is deployed as an ECS task, running the Streamlit application in a containerized environment.

## Detailed Steps

1. **Docker Image Build & Push**:
   - The CI/CD pipeline builds the Streamlit application into a Docker image.
   - The image is tagged with the commit SHA for version control.
   - The image is pushed to Amazon ECR for storage.

2. **Terraform-Based Infrastructure Deployment**:
   - The Terraform configuration provisions:
     - An ECR repository for storing images.
     - An ECS cluster with task definitions.
     - Security groups, IAM roles, and networking components.
   - The infrastructure is applied using GitLab CI/CD with `terraform apply`.

3. **Deploying the Application to ECS**:
   - The ECS task definition is updated with the latest image.
   - The ECS service is restarted to apply the new task definition.
   - Load balancing and auto-scaling ensure availability and scalability.

## ARCHITECTURE

In this architecture:

- The frontend container runs on Amazon ECS with Fargate.
- The container image is stored and retrieved from Amazon ECR.
- Network access is managed through security groups and, optionally, an Application Load Balancer.
- Terraform automates AWS infrastructure provisioning.

This setup ensures an automated, scalable, and reliable deployment pipeline for the frontend application.

![Architecture](<docs/img/DocuNinja Frontend-v0.drawio.png>)

## DEPLOYEMENT

### Requirements

The [saeiagendocuref infrastructure](https://git.cloud.safran/safranae/saeawspfiagendocuref/saeiagendocuref/frontend-chatbot/infrastructure/-/tree/main) should be deployed in your AWS account
up structure for tf 2025-09-19 10:09:31 +02:00			`# RAG Pattern :: Backend`
renommage de lib, rendre buildable 2025-08-08 10:57:00 +02:00
up structure for tf 2025-09-19 10:09:31 +02:00			`This repository hosts the code for the frontend application built with Streamlit. The deployment process involves building a Docker image, pushing it to Amazon Elastic Container Registry (ECR), and running it as an ECS task using Terraform for infrastructure management.`

			`## Deployment Workflow`

			`- :one: Docker Image Build and Push to ECR: The frontend application is containerized using Docker. The GitLab CI/CD pipeline builds the image and pushes it to an Amazon ECR repository.`
			`- :two: Infrastructure Provisioning with Terraform: The necessary AWS infrastructure, including the ECR repository, ECS cluster, and networking resources, is defined and managed using Terraform.`
			`- :three: ECS Task Deployment: The Docker image from ECR is deployed as an ECS task, running the Streamlit application in a containerized environment.`

			`## Detailed Steps`

			`1. Docker Image Build & Push:`
			`- The CI/CD pipeline builds the Streamlit application into a Docker image.`
			`- The image is tagged with the commit SHA for version control.`
			`- The image is pushed to Amazon ECR for storage.`

			`2. Terraform-Based Infrastructure Deployment:`
			`- The Terraform configuration provisions:`
			`- An ECR repository for storing images.`
			`- An ECS cluster with task definitions.`
			`- Security groups, IAM roles, and networking components.`
			- The infrastructure is applied using GitLab CI/CD with `terraform apply`.

			`3. Deploying the Application to ECS:`
			`- The ECS task definition is updated with the latest image.`
			`- The ECS service is restarted to apply the new task definition.`
			`- Load balancing and auto-scaling ensure availability and scalability.`

			`## ARCHITECTURE`

			`In this architecture:`

			`- The frontend container runs on Amazon ECS with Fargate.`
			`- The container image is stored and retrieved from Amazon ECR.`
			`- Network access is managed through security groups and, optionally, an Application Load Balancer.`
			`- Terraform automates AWS infrastructure provisioning.`

			`This setup ensures an automated, scalable, and reliable deployment pipeline for the frontend application.`

			`![Architecture](<docs/img/DocuNinja Frontend-v0.drawio.png>)`

			`## DEPLOYEMENT`

			`### Requirements`

			`The [saeiagendocuref infrastructure](https://git.cloud.safran/safranae/saeawspfiagendocuref/saeiagendocuref/frontend-chatbot/infrastructure/-/tree/main) should be deployed in your AWS account`