diff --git a/platforms/gke-aiml/playground/README.md b/platforms/gke-aiml/playground/README.md
index 893750e5..34251dfb 100644
--- a/platforms/gke-aiml/playground/README.md
+++ b/platforms/gke-aiml/playground/README.md
@@ -21,13 +21,9 @@ document.
 
 ### Project
 
-<<<<<<< HEAD In this guide you can choose to bring your project (BYOP) or have
-Terraform create a new project for you. The requirements are difference based on
-the option that you choose. ======= In this guide you can choose to bring your
-project (BYOP) or have Terraform create a new project for you. The requirements
-are different based on the option that you choose.
-
-> > > > > > > 04b228a (Added AlloyDB with PSC to the playground (#36))
+In this guide you can choose to bring your project (BYOP) or have Terraform
+create a new project for you. The requirements are different based on the option
+that you choose.
 
 #### Option 1: Bring your own project (BYOP)
 
@@ -339,7 +335,8 @@ Management API are enabled.
   gcloud endpoints services undelete gradio.ml-team.mlp-${MLP_ENVIRONMENT_NAME}.endpoints.${MLP_PROJECT_ID}.cloud.goog --quiet 2>/dev/null
   gcloud endpoints services undelete locust.ml-team.mlp-${MLP_ENVIRONMENT_NAME}.endpoints.${MLP_PROJECT_ID}.cloud.goog --quiet 2>/dev/null
   gcloud endpoints services undelete mlflow-tracking.ml-team.mlp-${MLP_ENVIRONMENT_NAME}.endpoints.${MLP_PROJECT_ID}.cloud.goog --quiet 2>/dev/null
-  gcloud endpoints services undelete ray-dashboard.ml-team.mlp-${MLP_ENVIRONMENT_NAME}.endpoints.${MLP_PROJECT_ID}.cloud.goog --quiet 2>/dev/null
+  gcloud endpoints services undelete rag-frontend.ml-team.mlp-${MLP_ENVIRONMENT_NAME}.endpoints.${MLP_PROJECT_ID}.cloud.goog --quiet 2>/dev/null
+  gcloud endpoints services undelete ray-dashboard.ml-team.mlp-${MLP_ENVIRONMENT_NAME}.endpoints.${MLP_PROJECT_ID}.cloud.goog --quiet 2>/dev/null1
   ```
 
 - Create the resources
diff --git a/use-cases/rag-pipeline/README.md b/use-cases/rag-pipeline/README.md
index b35651fa..24c74c1d 100644
--- a/use-cases/rag-pipeline/README.md
+++ b/use-cases/rag-pipeline/README.md
@@ -9,7 +9,7 @@ the catalog. This approach not only improves the customer experience by
 providing relevant alternatives but also helps reduce lost sales and potentially
 increases average order value.
 
-## Here's how it works:
+## How it works
 
 Understanding Customer Intent: The system analyzes the customer's search query
 (e.g., "blue cotton t-shirt") to understand the key attributes and features they
@@ -24,7 +24,7 @@ even if they don't share the exact same words.
 system generates a list of relevant product recommendations. These might
 include:
 
-**Cosine Simlarity:** Cosine similarity measures how similar two items are by
+**Cosine Similarity:** Cosine similarity measures how similar two items are by
 looking at the angle between their vector representations.
 
 Cosine: Cosine is a mathematical function that tells you how much two vectors
@@ -55,14 +55,6 @@ sweaters.
 relevance and presented to the customer in a clear and appealing way,
 encouraging them to continue shopping.
 
-## Data Preprocessing for RAG
-
-We need a input dataset to feed to our RAG pipeline. We take a raw dataset and
-filter and clean it up to prepare it for our RAG pipeline. Perform the data
-preprocessing steps as described in
-[README](/use-cases/rag-pipeline/data-preprocessing/README.md) to prepare input
-dataset.
-
 ## Architecture
 
 ![RAG Architecture](./docs/arch-rag-architecture-flow.png)
@@ -83,8 +75,8 @@ Let's break down the flow step-by-step:
 
 2. **Get Embeddings:**
 
-   The user's query is sent to the "Backend Application". Within the Backend
-   Application, the query is processed by an "Embedding Model Endpoint".
+   The user's query is sent to the "backend application". Within the backend
+   application, the query is processed by an "embedding model endpoint".
 
 3. **Embedding Model Endpoint:**
 
@@ -96,8 +88,8 @@ Let's break down the flow step-by-step:
 
 4. **Get Embeddings (Continued):**
 
-   The Backend Application receives the embedding vector from the Embedding
-   Model Endpoint.
+   The backend application receives the embedding vector from the embedding
+   model endpoint.
 
 5. **Scann Index:**
 
@@ -121,12 +113,12 @@ Let's break down the flow step-by-step:
 7. **Instruction Tuned Model Endpoint:**
 
    These retrieved similar product information is then sent to an "Instruction
-   Tuned Model Endpoint". This endpoint hosts a specific version of Gemini
+   Tuned Model Endpoint". This endpoint hosts a specific version of the Gemma 2B
    instruction tuned
    model(gemma-2b-it)[https://huggingface.co/google/gemma-2b-it] that's been
    trained with a focus on understanding and responding to instructions
-   effectively. Instruction tuned model is provided with a specific instructions
-   as a prompt to re-rank the search results.
+   effectively. The instruction tuned model is provided with a specific
+   instructions as a prompt to re-rank the search results.
 
 8. **Re-Rank Search Results with LLM:**
 
@@ -136,7 +128,7 @@ Let's break down the flow step-by-step:
 
 9. **Suggest Products:**
 
-   The Backend Application receives the re-ranked product list from the
+   The backend application receives the re-ranked product list from the
    Instruction Tuned Model Endpoint.This list is sent as Product Recommendations
    to the Gradio Chat Interface.
 
@@ -156,99 +148,74 @@ the Scann Index to reflect catalog changes
 This section outlines the steps to set up the Retrieval Augmented Generation
 (RAG) pipeline for product recommendations.
 
-1. **Create the Vector Store:** Create the `product_catalog` database in
-   [AlloyDB](https://cloud.google.com/alloydb/docs/introduction). This database
-   will house the `clothes` table, which stores product catalog information.
-
-2. **Deploy the Embedding Model:** Deploy the
+1. **Deploy the Embedding Model:** Deploy the
    [Blip2 multimodal](https://github.com/salesforce/LAVIS/blob/main/examples/blip_feature_extraction.ipynb)embedding
    model. This model generates text, image, and multimodal embeddings for each
    product.
 
-3. **Generate and Store Embeddings:** Use an ETL pipeline to generate embeddings
+1. **Create the Vector Store:** Create the `product_catalog` database in
+   [AlloyDB](https://cloud.google.com/alloydb/docs/introduction). This database
+   will house the `clothes` table, which stores product catalog information.
+
+1. **Generate and Store Embeddings:** Use an ETL pipeline to generate embeddings
    (text, image, and multimodal) using the deployed Blip2 model. Store these
    embeddings in separate columns within the `clothes` table in AlloyDB.
 
-4. **Deploy the Instruction-Tuned Model:** Deploy the
+1. **Deploy the Instruction-Tuned Model:** Deploy the
    [gemma-2b-it model](https://huggingface.co/google/gemma-2b-it). This model
    generates natural language responses and product recommendations based on
    user queries and retrieved product information.
 
-5. **Deploy the Backend API:** Deploy the FastAPI backend. This API serves as
+1. **Deploy the Backend API:** Deploy the FastAPI backend. This API serves as
    the interface between the user interface, embedding model, instruction-tuned
    model, and the AlloyDB vector store. It processes user prompts and generates
    product recommendations.
 
-6. **Deploy the Frontend UI:** Deploy the [gradio](https://gradio.app/) based
+1. **Deploy the Frontend UI:** Deploy the [gradio](https://gradio.app/) based
    frontend UI. This UI provides a chatbot interface for end-users to interact
    with the RAG pipeline and receive product recommendations.
 
-## Prerequisites
+## Deploy RAG Components
+
+This section outlines the steps to deploy the Retrieval Augmented Generation
+(RAG) pipeline to the playground cluster.
+
+### Prerequisites
 
 - Use the existing
   [playground AI/ML platform](/platforms/gke-aiml/playground/README.md). If you
   are using a different environment the scripts and manifest will need to be
   modified for that environment.
-- Run [data preprocessing for RAG](#data-preprocessing-for-rag)
-
-#### Set variable for the ML playground environment
-
-- Clone the repository
-
-  ```sh
-  git clone https://github.com/GoogleCloudPlatform/accelerated-platforms && \
-  cd accelerated-platforms
-  ```
-
-- Change directory to the guide directory
-
-  ```sh
-  cd use-cases/rag-pipeline
-  ```
-
-- Ensure that your `MLP_ENVIRONMENT_FILE` is configured
-
-  ```sh
-  cat ${MLP_ENVIRONMENT_FILE} && \
-  source ${MLP_ENVIRONMENT_FILE}
-  ```
-
-  > You should see the various variables populated with the information specific
-  > to your environment.
 
-- Get credentials for the GKE cluster
+### Data Preprocessing for RAG
 
-  ```sh
-  gcloud container fleet memberships get-credentials ${MLP_CLUSTER_NAME} --project ${MLP_PROJECT_ID}
-  ```
-
-# Deploy RAG Application Components
-
-This section outlines the steps to deploy the Retrieval Augmented Generation
-(RAG) pipeline to the playground cluster. The components should be deployed in
-the following order:
+We need a input dataset to feed to our RAG pipeline. We take a raw dataset and
+filter and clean it up to prepare it for our RAG pipeline. Perform the data
+preprocessing steps as described in
+[README](/use-cases/rag-pipeline/data-preprocessing/README.md) to prepare input
+dataset.
 
-## Deploy the Multimodal Model on the playground cluster
+### Deploy the Multimodal Model on the playground cluster
 
 Deploy multimodal model on ML playground, follow the
 [README](/use-cases/rag-pipeline/embedding-models/multimodal-embedding/README.md)
 
-## Deploy instruction tuned model on the playground cluster
+### Deploy instruction tuned model on the playground cluster
 
 Deploy instruction tuned model on ML playground, follow the
 [README](/use-cases/rag-pipeline/instruction-tuned-model/README.md)
 
-## Create database `product_catalog` in alloyDB to import Product Catalog
+### Create database `product_catalog` in alloyDB to import Product Catalog
 
 Deploy database setup kubernetes job on the ML playground cluster, follow the
 [README](/use-cases/rag-pipeline/alloy-db-setup/README.md)
 
-## Deploy the backend on the playground cluster
+### Deploy the backend on the playground cluster
 
 Deploy backend application on the ML playground cluster, follow the
 [README](/use-cases/rag-pipeline/backend/README.md)
 
-## Deploy the frontend on the playground cluster
+### Deploy the frontend on the playground cluster
 
 Deploy frontend application on the MLP playground cluster, follow the
 [README](/use-cases/rag-pipeline/frontend/README.md)
diff --git a/use-cases/rag-pipeline/alloy-db-setup/README.md b/use-cases/rag-pipeline/alloy-db-setup/README.md
index 57d68931..5c050ef3 100644
--- a/use-cases/rag-pipeline/alloy-db-setup/README.md
+++ b/use-cases/rag-pipeline/alloy-db-setup/README.md
@@ -1,4 +1,4 @@
-# Process to set up AlloyDB
+# RAG: Database setup and initialization
 
 This kubernetes job helps you load the flipkart product catalog to the alloyDB
 database named `product_catalog`.Also it creates separate columns to store the
@@ -7,41 +7,31 @@ embeddings(text, image and multimodal) in a table named `clothes` in the
 
 ## Prerequisites
 
-<TODO> Write few lines about alloydb set up various users for IAM, workload
-identity , different users in ML_ENV_FILE to use .
-
-<TODO> Decide what how end users should get access to these buckets of the
-image_uris and the associated datasets to load the product catalog?
-
-MLP accounts MLP_DB_ADMIN_IAM and MLP_DB_USER_IAM need Storage object
-permissions to retrieve, process and generate embeddings for image_uri stores in
-Cloud Storage buckets.
-
 - This guide was developed to be run on the
   [playground AI/ML platform](/platforms/gke-aiml/playground/README.md). If you
   are using a different environment the scripts and manifest will need to be
   modified for that environment.
-- Multimodal embedding model has been deployed as per instructions in the
-  embedding models folder (../embedding-models/README.md)
+- [RAG: Multimodal embedding model](/use-cases/rag-pipeline/embedding-models/multimodal-embedding/README.md)
+  has been deployed as per the instructions.
 
 ## Preparation
 
-- Clone the repository
+- Clone the repository.
 
-  ```sh
+  ```shell
   git clone https://github.com/GoogleCloudPlatform/accelerated-platforms && \
   cd accelerated-platforms
   ```
 
-- Change directory to the guide directory
+- Change directory to the guide directory.
 
-  ```sh
+  ```shell
   cd use-cases/rag-pipeline/alloy-db-setup
   ```
 
-- Ensure that your `MLP_ENVIRONMENT_FILE` is configured
+- Ensure that your `MLP_ENVIRONMENT_FILE` is configured.
 
-  ```sh
+  ```shell
   cat ${MLP_ENVIRONMENT_FILE} && \
   set -o allexport && \
   source ${MLP_ENVIRONMENT_FILE} && \
@@ -51,10 +41,13 @@ Cloud Storage buckets.
   > You should see the various variables populated with the information specific
   > to your environment.
 
-- Get credentials for the GKE cluster
+- Get credentials for the GKE cluster.
 
-  ```sh
-  gcloud container fleet memberships get-credentials ${MLP_CLUSTER_NAME} --project ${MLP_PROJECT_ID}
+  ```shell
+  gcloud container clusters get-credentials ${MLP_CLUSTER_NAME} \
+  --dns-endpoint \
+  --project=${MLP_PROJECT_ID} \
+  --region=${MLP_REGION}
   ```
 
 ## Build the container image
@@ -62,7 +55,7 @@ Cloud Storage buckets.
 - Build the container image using Cloud Build and push the image to Artifact
   Registry
 
-  ```sh
+  ```shell
   cd src
   git restore cloudbuild.yaml
   sed -i -e "s|^serviceAccount:.*|serviceAccount: projects/${MLP_PROJECT_ID}/serviceAccounts/${MLP_BUILD_GSA}|" cloudbuild.yaml
@@ -79,22 +72,9 @@ Cloud Storage buckets.
 
 ## Run the job
 
-**Steps to produce the `MASTER_CATALOG_FILE_NAME` need to be included
-somewhere**
-
-- Temporary steps to populate required data
-
-  ```
-  gcloud storage cp gs://temporary-rag-data/master_product_catalog.csv . && \
-  sed -i s"/<MLP_DATA_BUCKET>/${MLP_DATA_BUCKET}/g" master_product_catalog.csv && \
-  gcloud storage cp master_product_catalog.csv gs://${MLP_DATA_BUCKET}/ && \
-  rm -f master_product_catalog.csv && \
-  gcloud storage rsync gs://temporary-rag-data/flipkart_images gs://${MLP_DATA_BUCKET}/flipkart_images/
-  ```
-
 - Configure the job
 
-  ```sh
+  ```shell
   set -o nounset
   export CATALOG_DB_NAME="product_catalog"
   export CATALOG_TABLE_NAME="clothes"
@@ -107,12 +87,15 @@ somewhere**
   export EMBEDDING_ENDPOINT_IMAGE="http://multimodal-embedding-model.ml-team:80/image_embeddings"
   export EMBEDDING_ENDPOINT_MULTIMODAL="http://multimodal-embedding-model.ml-team:80/multimodal_embeddings"
   export EMBEDDING_ENDPOINT_TEXT="http://multimodal-embedding-model.ml-team:80/text_embeddings"
-  export MASTER_CATALOG_FILE_NAME="master_product_catalog.csv"
+  export MASTER_CATALOG_FILE_NAME="RAG/master_product_catalog.csv"
   export NUM_LEAVES_VALUE="300"
   set +o nounset
   ```
 
-  ```sh
+  > Ensure there are no `bash: <ENVIRONMENT_VARIABLE> unbound variable` error
+  > messages.
+
+  ```shell
   git restore manifests/job-initialize-database.yaml manifests/job-populate-table.yaml
   envsubst < manifests/job-initialize-database.yaml | sponge manifests/job-initialize-database.yaml
   envsubst < manifests/job-populate-table.yaml | sponge manifests/job-populate-table.yaml
@@ -120,7 +103,7 @@ somewhere**
 
 - Create the initialize database job.
 
-  ```
+  ```shell
   kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} apply -f manifests/job-initialize-database.yaml
   ```
 
@@ -128,22 +111,27 @@ somewhere**
 
 - Watch the job until it is complete.
 
-  ```
+  ```shell
   watch --color --interval 5 --no-title \
   "kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} get job/initialize-database | GREP_COLORS='mt=01;92' egrep --color=always -e '^' -e 'Complete'
   echo '\nLogs(last 10 lines):'
   kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} logs job/initialize-database --tail 10"
   ```
 
+  ```
+  NAME                  STATUS     COMPLETIONS   DURATION   AGE
+  initialize-database   Complete   1/1           XXXXX      XXXXX
+  ```
+
 - Check logs for any errors.
 
-  ```
+  ```shell
   kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} logs job/initialize-database
   ```
 
 - Create the populate table job.
 
-  ```
+  ```shell
   kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} apply -f manifests/job-populate-table.yaml
   ```
 
@@ -151,15 +139,20 @@ somewhere**
 
 - Watch the job until it is complete.
 
-  ```
+  ```shell
   watch --color --interval 5 --no-title \
   "kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} get job/populate-table | GREP_COLORS='mt=01;92' egrep --color=always -e '^' -e 'Complete'
   echo '\nLogs(last 10 lines):'
   kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} logs job/populate-table --tail 10"
   ```
 
+  ```
+  NAME             STATUS     COMPLETIONS   DURATION   AGE
+  populate-table   Complete   1/1           XXXXX      XXXXX
+  ```
+
 - Check logs for any errors.
 
-  ```
+  ```shell
   kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} logs job/populate-table
   ```
diff --git a/use-cases/rag-pipeline/backend/README.md b/use-cases/rag-pipeline/backend/README.md
index e03b4eff..d7af432a 100644
--- a/use-cases/rag-pipeline/backend/README.md
+++ b/use-cases/rag-pipeline/backend/README.md
@@ -1,4 +1,4 @@
-# Backend application deployment
+# RAG: Backend deployment
 
 ## Prerequisites
 
@@ -9,41 +9,46 @@
 
 ## Preparation
 
-- Clone the repository
+- Clone the repository.
 
-  ```sh
+  ```shell
   git clone https://github.com/GoogleCloudPlatform/accelerated-platforms && \
   cd accelerated-platforms
   ```
 
-- Change directory to the guide directory
+- Change directory to the guide directory.
 
-  ```sh
+  ```shell
   cd use-cases/rag-pipeline/backend
   ```
 
-- Ensure that your `MLP_ENVIRONMENT_FILE` is configured
+- Ensure that your `MLP_ENVIRONMENT_FILE` is configured.
 
-  ```sh
+  ```shell
   cat ${MLP_ENVIRONMENT_FILE} && \
-  source ${MLP_ENVIRONMENT_FILE}
+  set -o allexport && \
+  source ${MLP_ENVIRONMENT_FILE} && \
+  set +o allexport
   ```
 
   > You should see the various variables populated with the information specific
   > to your environment.
 
-- Get credentials for the GKE cluster
+- Get credentials for the GKE cluster.
 
-  ```sh
-  gcloud container fleet memberships get-credentials ${MLP_CLUSTER_NAME} --project ${MLP_PROJECT_ID}
+  ```shell
+  gcloud container clusters get-credentials ${MLP_CLUSTER_NAME} \
+  --dns-endpoint \
+  --project=${MLP_PROJECT_ID} \
+  --region=${MLP_REGION}
   ```
 
 ## Build the container image
 
 - Build the container image using Cloud Build and push the image to Artifact
-  Registry
+  Registry.
 
-  ```sh
+  ```shell
   cd src
   git restore cloudbuild.yaml
   sed -i -e "s|^serviceAccount:.*|serviceAccount: projects/${MLP_PROJECT_ID}/serviceAccounts/${MLP_BUILD_GSA}|" cloudbuild.yaml
@@ -55,67 +60,67 @@
   cd -
   ```
 
-## Deploy the backend application
+## Deploy the backend
 
 - Configure the deployment.
 
-  ```sh
+  ```shell
+  set -o nounset
   export CATALOG_DB="product_catalog"
   export CATALOG_TABLE_NAME="clothes"
-  export TEXT_EMBEDDING_ENDPOINT="http://multimodal-embedding-model.ml-team:80/text_embeddings"
-  export IMAGE_EMBEDDING_ENDPOINT="http://multimodal-embedding-model.ml-team:80/image_embeddings"
-  export MULTIMODAL_EMBEDDING_ENDPOINT="http://multimodal-embedding-model.ml-team:80/multimodal_embeddings"
-  export GEMMA_IT_ENDPOINT="http://rag-it-model.ml-team:8000/v1/chat/completions"
-  export EMBEDDING_COLUMN_TEXT="text_embeddings"
+  export CONTAINER_IMAGE_URL="${MLP_RAG_BACKEND_IMAGE}"
+  export DB_INSTANCE_URI="${MLP_DB_INSTANCE_URI}"
   export EMBEDDING_COLUMN_IMAGE="image_embeddings"
   export EMBEDDING_COLUMN_MULTIMODAL="multimodal_embeddings"
-  export ROW_COUNT="\"5\""
+  export EMBEDDING_COLUMN_TEXT="text_embeddings"
+  export EMBEDDING_ENDPOINT_IMAGE="http://multimodal-embedding-model.ml-team:80/image_embeddings"
+  export EMBEDDING_ENDPOINT_MULTIMODAL="http://multimodal-embedding-model.ml-team:80/multimodal_embeddings"
+  export EMBEDDING_ENDPOINT_TEXT="http://multimodal-embedding-model.ml-team:80/text_embeddings"
+  export GEMMA_IT_ENDPOINT="http://rag-it-model.ml-team:8000/v1/chat/completions"
+  export KUBERNETES_SERVICE_ACCOUNT="${MLP_DB_USER_KSA}"
+
+  export ROW_COUNT="5"
+  set +o nounset
   ```
 
-  ```sh
+  > Ensure there are no `bash: <ENVIRONMENT_VARIABLE> unbound variable` error
+  > messages.
+
+  ```shell
   git restore manifests/deployment.yaml
-  sed \
-  -i -e "s|V_CATALOG_DB|${CATALOG_DB}|" \
-  -i -e "s|V_CATALOG_TABLE_NAME|${CATALOG_TABLE_NAME}|" \
-  -i -e "s|V_EMBEDDING_COLUMN_IMAGE|${EMBEDDING_COLUMN_IMAGE}|" \
-  -i -e "s|V_EMBEDDING_COLUMN_MULTIMODAL|${EMBEDDING_COLUMN_MULTIMODAL}|" \
-  -i -e "s|V_EMBEDDING_COLUMN_TEXT|${EMBEDDING_COLUMN_TEXT}|" \
-  -i -e "s|V_EMBEDDING_ENDPOINT_IMAGE|${IMAGE_EMBEDDING_ENDPOINT}|" \
-  -i -e "s|V_EMBEDDING_ENDPOINT_MULTIMODAL|${MULTIMODAL_EMBEDDING_ENDPOINT}|" \
-  -i -e "s|V_EMBEDDING_ENDPOINT_TEXT|${TEXT_EMBEDDING_ENDPOINT}|" \
-  -i -e "s|V_GEMMA_IT_ENDPOINT|${GEMMA_IT_ENDPOINT}|" \
-  -i -e "s|V_IMAGE|${MLP_RAG_BACKEND_IMAGE}|" \
-  -i -e "s|V_MLP_DB_ADMIN_IAM|${MLP_DB_ADMIN_IAM}|" \
-  -i -e "s|V_MLP_DB_USER_KSA|${MLP_DB_USER_KSA}|" \
-  -i -e "s|V_MLP_DB_INSTANCE_URI|${MLP_DB_INSTANCE_URI}|" \
-  -i -e "s|V_MLP_KUBERNETES_NAMESPACE|${MLP_KUBERNETES_NAMESPACE}|" \
-  -i -e "s|V_PROJECT_ID|${MLP_PROJECT_ID}|" \
-  -i -e "s|V_ROW_COUNT|${ROW_COUNT}|" \
-  manifests/deployment.yaml
+  envsubst < manifests/deployment.yaml | sponge manifests/deployment.yaml
   ```
 
 - Create the deployment.
 
-  ```sh
+  ```shell
   kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} apply -f manifests/deployment.yaml
   ```
 
-## Test the backend application
+- Watch the deployment until it is ready and available.
 
-Validations:
+  ```shell
+  watch --color --interval 5 --no-title \
+  "kubectl --namespace ${MLP_MODEL_OPS_NAMESPACE} get deployment/rag-backend | GREP_COLORS='mt=01;92' egrep --color=always -e '^' -e '1/1     1            1'
+  echo '\nLogs(last 10 lines):'
+  kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} logs deployment/rag-backend --tail 10"
+  ```
 
-```sh
-kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} get pods -l app=rag-backend
-```
+  ```
+  NAME          READY   UP-TO-DATE   AVAILABLE   AGE
+  rag-backend   1/1     1            1           XXXXX
+  ```
+
+## Verify the backend
 
-```sh
-kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} get service/rag-backend
-```
+- Create the curl job.
 
-```sh
-kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} apply -f manifests/curl.yaml
-```
+  ```shell
+  kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} apply -f manifests/curl.yaml
+  ```
+
+- Get the logs for the curl job.
 
-```sh
-kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} logs job/rag-backend-curl
-```
+  ```shell
+  kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} logs --follow job/rag-backend-curl
+  ```
diff --git a/use-cases/rag-pipeline/backend/manifests/deployment.yaml b/use-cases/rag-pipeline/backend/manifests/deployment.yaml
index 3aaa633d..0a183335 100644
--- a/use-cases/rag-pipeline/backend/manifests/deployment.yaml
+++ b/use-cases/rag-pipeline/backend/manifests/deployment.yaml
@@ -26,38 +26,36 @@ spec:
       labels:
         app: rag-backend
     spec:
-      serviceAccountName: V_MLP_DB_USER_KSA
+      serviceAccountName: ${KUBERNETES_SERVICE_ACCOUNT}
       containers:
       - name: rag
-        image: V_IMAGE
+        image: ${CONTAINER_IMAGE_URL}
         imagePullPolicy: Always
         env:
         - name: CATALOG_DB
-          value: V_CATALOG_DB
+          value: "${CATALOG_DB}"
         - name: CATALOG_TABLE_NAME
-          value: V_CATALOG_TABLE_NAME
-        - name: MLP_DB_ADMIN_IAM
-          value: V_MLP_DB_ADMIN_IAM
+          value: "${CATALOG_TABLE_NAME}"
         - name: MLP_DB_INSTANCE_URI
-          value: V_MLP_DB_INSTANCE_URI
+          value: "${DB_INSTANCE_URI}"
         - name: GEMMA_IT_ENDPOINT
-          value: V_GEMMA_IT_ENDPOINT
+          value: "${GEMMA_IT_ENDPOINT}"
         - name: MLP_KUBERNETES_NAMESPACE
-          value: V_MLP_KUBERNETES_NAMESPACE
+          value: "${MLP_KUBERNETES_NAMESPACE}"
         - name: TEXT_EMBEDDING_ENDPOINT
-          value: V_EMBEDDING_ENDPOINT_TEXT
+          value: "${EMBEDDING_ENDPOINT_TEXT}"
         - name: IMAGE_EMBEDDING_ENDPOINT
-          value: V_EMBEDDING_ENDPOINT_IMAGE
+          value: "${EMBEDDING_ENDPOINT_IMAGE}"
         - name: MULTIMODAL_EMBEDDING_ENDPOINT
-          value: V_EMBEDDING_ENDPOINT_MULTIMODAL
+          value: "${EMBEDDING_ENDPOINT_MULTIMODAL}"
         - name: EMBEDDING_COLUMN_TEXT
-          value: V_EMBEDDING_COLUMN_TEXT
+          value: "${EMBEDDING_COLUMN_TEXT}"
         - name: EMBEDDING_COLUMN_IMAGE
-          value: V_EMBEDDING_COLUMN_IMAGE
+          value: "${EMBEDDING_COLUMN_IMAGE}"
         - name: EMBEDDING_COLUMN_MULTIMODAL
-          value: V_EMBEDDING_COLUMN_MULTIMODAL
+          value: "${EMBEDDING_COLUMN_MULTIMODAL}"
         - name: ROW_COUNT
-          value: V_ROW_COUNT
+          value: "${ROW_COUNT}"
         resources:
           requests:
             cpu: "2"
diff --git a/use-cases/rag-pipeline/data-preprocessing/README.md b/use-cases/rag-pipeline/data-preprocessing/README.md
index 451938a4..6832e526 100644
--- a/use-cases/rag-pipeline/data-preprocessing/README.md
+++ b/use-cases/rag-pipeline/data-preprocessing/README.md
@@ -1,4 +1,4 @@
-# Data Preprocessing for RAG
+# RAG: Data preprocessing
 
 ## Dataset
 
@@ -36,66 +36,81 @@ The data preprocessing step takes approximately 18-20 minutes.
 
 ## Preparation
 
-- Clone the repository
+- Clone the repository.
 
   ```shell
   git clone https://github.com/GoogleCloudPlatform/accelerated-platforms && \
   cd accelerated-platforms
   ```
 
-- Change directory to the guide directory
+- Change directory to the guide directory.
 
   ```shell
   cd use-cases/rag-pipeline/data-preprocessing
   ```
 
-- Ensure that your `MLP_ENVIRONMENT_FILE` is configured
+- Ensure that your `MLP_ENVIRONMENT_FILE` is configured.
 
   ```shell
   cat ${MLP_ENVIRONMENT_FILE} && \
-  source ${MLP_ENVIRONMENT_FILE}
+  set -o allexport && \
+  source ${MLP_ENVIRONMENT_FILE} && \
+  set +o allexport
   ```
 
   > You should see the various variables populated with the information specific
   > to your environment.
 
+- Get credentials for the GKE cluster.
+
+  ```shell
+  gcloud container clusters get-credentials ${MLP_CLUSTER_NAME} \
+  --dns-endpoint \
+  --project=${MLP_PROJECT_ID} \
+  --region=${MLP_REGION}
+  ```
+
 ## Build the container image
 
 - Build container image using Cloud Build and push the image to Artifact
-  Registry
+  Registry.
 
   ```shell
   cd src
+  rm -rf datapreprocessing
   cp -r ${MLP_BASE_DIR}/modules/python/src/datapreprocessing .
+  git restore cloudbuild.yaml
   sed -i -e "s|^serviceAccount:.*|serviceAccount: projects/${MLP_PROJECT_ID}/serviceAccounts/${MLP_BUILD_GSA}|" cloudbuild.yaml
   gcloud beta builds submit \
   --config cloudbuild.yaml \
   --gcs-source-staging-dir gs://${MLP_CLOUDBUILD_BUCKET}/source \
   --project ${MLP_PROJECT_ID} \
   --substitutions _DESTINATION=${MLP_RAG_DATA_PROCESSING_IMAGE}
-  rm -rf datapreprocessing
   cd ..
   ```
 
 ## Run the job
 
-- Get credentials for the GKE cluster
+- Configure the job.
 
   ```shell
-  gcloud container fleet memberships get-credentials ${MLP_CLUSTER_NAME} --project ${MLP_PROJECT_ID}
+  set -o nounset
+  export CONTAINER_IMAGE_URL="${MLP_RAG_DATA_PROCESSING_IMAGE}"
+  export DATA_BUCKET="${MLP_DATA_BUCKET}"
+  export KUBERNETES_SERVICE_ACCOUNT="${MLP_RAG_DATA_PROCESSING_KSA}"
+  export RAY_CLUSTER_HOST="ray-cluster-kuberay-head-svc.ml-team:10001"
+  set +o nounset
   ```
 
-- Configure the job
+  > Ensure there are no `bash: <ENVIRONMENT_VARIABLE> unbound variable` error
+  > messages.
 
   ```shell
-  sed \
-  -i -e "s|V_DATA_BUCKET|${MLP_DATA_BUCKET}|" \
-  -i -e "s|V_IMAGE_URL|${MLP_RAG_DATA_PROCESSING_IMAGE}|" \
-  -i -e "s|V_KSA|${MLP_RAG_DATA_PROCESSING_KSA}|" \
-  manifests/job.yaml
+  git restore manifests/job.yaml
+  envsubst < manifests/job.yaml | sponge manifests/job.yaml
   ```
 
-- Create the job
+- Create the job.
 
   ```shell
   kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} apply -f manifests/job.yaml
diff --git a/use-cases/rag-pipeline/data-preprocessing/manifests/job.yaml b/use-cases/rag-pipeline/data-preprocessing/manifests/job.yaml
index 7982bc1a..d6ecc234 100644
--- a/use-cases/rag-pipeline/data-preprocessing/manifests/job.yaml
+++ b/use-cases/rag-pipeline/data-preprocessing/manifests/job.yaml
@@ -26,13 +26,13 @@ spec:
     spec:
       containers:
       - name: job
-        image: V_IMAGE_URL
+        image: ${CONTAINER_IMAGE_URL}
         imagePullPolicy: Always
         env:
         - name: "PROCESSING_BUCKET"
-          value: V_DATA_BUCKET
+          value: "${DATA_BUCKET}"
         - name: "RAY_CLUSTER_HOST"
-          value: ray-cluster-kuberay-head-svc.ml-team:10001
+          value: "${RAY_CLUSTER_HOST}"
         resources:
           requests:
             cpu: 100m
@@ -43,7 +43,7 @@ spec:
       nodeSelector:
         resource-type: cpu
       restartPolicy: Never
-      serviceAccountName: V_KSA
+      serviceAccountName: ${KUBERNETES_SERVICE_ACCOUNT}
       tolerations:
       - effect: NoSchedule
         key: on-demand
diff --git a/use-cases/rag-pipeline/embedding-models/multimodal-embedding/README.md b/use-cases/rag-pipeline/embedding-models/multimodal-embedding/README.md
index 99c747b1..8284a6ac 100644
--- a/use-cases/rag-pipeline/embedding-models/multimodal-embedding/README.md
+++ b/use-cases/rag-pipeline/embedding-models/multimodal-embedding/README.md
@@ -1,4 +1,4 @@
-# Multimodal blip2 model
+# RAG: Multimodal embedding model
 
 To know more about the embedding model see original
 [blog](https://blog.salesforceairesearch.com/blip-2/) and
@@ -13,41 +13,46 @@ To know more about the embedding model see original
 
 ## Preparation
 
-- Clone the repository
+- Clone the repository.
 
-  ```sh
+  ```shell
   git clone https://github.com/GoogleCloudPlatform/accelerated-platforms && \
   cd accelerated-platforms
   ```
 
-- Change directory to the guide directory
+- Change directory to the guide directory.
 
-  ```sh
+  ```shell
   cd use-cases/rag-pipeline/embedding-models/multimodal-embedding
   ```
 
-- Ensure that your `MLP_ENVIRONMENT_FILE` is configured
+- Ensure that your `MLP_ENVIRONMENT_FILE` is configured.
 
-  ```sh
+  ```shell
   cat ${MLP_ENVIRONMENT_FILE} && \
-  source ${MLP_ENVIRONMENT_FILE}
+  set -o allexport && \
+  source ${MLP_ENVIRONMENT_FILE} && \
+  set +o allexport
   ```
 
   > You should see the various variables populated with the information specific
   > to your environment.
 
-- Get credentials for the GKE cluster
+- Get credentials for the GKE cluster.
 
-  ```sh
-  gcloud container fleet memberships get-credentials ${MLP_CLUSTER_NAME} --project ${MLP_PROJECT_ID}
+  ```shell
+  gcloud container clusters get-credentials ${MLP_CLUSTER_NAME} \
+  --dns-endpoint \
+  --project=${MLP_PROJECT_ID} \
+  --region=${MLP_REGION}
   ```
 
 ## Build the container image
 
 - Build the container image using Cloud Build and push the image to Artifact
-  Registry
+  Registry.
 
-  ```sh
+  ```shell
   cd src
   git restore cloudbuild.yaml
   sed -i -e "s|^serviceAccount:.*|serviceAccount: projects/${MLP_PROJECT_ID}/serviceAccounts/${MLP_BUILD_GSA}|" cloudbuild.yaml
@@ -59,67 +64,73 @@ To know more about the embedding model see original
   cd -
   ```
 
-## Deploy the embedding model
+## Deploy the model
 
-- Configure the deployment
+- Configure the deployment.
 
-  ```sh
-  git restore manifests/deployment.yaml
-  sed \
-  -i -e "s|V_IMAGE|${MLP_MULTIMODAL_EMBEDDING_IMAGE}|" \
-  -i -e "s|V_KSA|${MLP_DB_USER_KSA}|" \
-  manifests/deployment.yaml
+  ```shell
+  set -o nounset
+  export CONTAINER_IMAGE_URL="${MLP_MULTIMODAL_EMBEDDING_IMAGE}"
+  export KUBERNETES_SERVICE_ACCOUNT="${MLP_DB_USER_KSA}"
+  set +o nounset
   ```
 
-- Create the deployment
+  > Ensure there are no `bash: <ENVIRONMENT_VARIABLE> unbound variable` error
+  > messages.
 
-  ```sh
-  kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} apply -f manifests/deployment.yaml
+  ```shell
+  git restore manifests/deployment.yaml
+  envsubst < manifests/deployment.yaml | sponge manifests/deployment.yaml
   ```
 
-## Validate the embedding model deployment
-
-```sh
-kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} get pods -l app=multimodal-embedding-model
-```
+- Create the deployment.
 
-```sh
-kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} get service/multimodal-embedding-model
-```
-
-## Verify the embedding model
+  ```shell
+  kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} apply -f manifests/deployment.yaml
+  ```
 
-**This assumes that the `flipkart_images` are already present in the
-`MLP_DATA_BUCKET`. If they are not, run the following command**
+- Watch the deployment until it is ready and available.
 
-- Temporary steps to populate required data
+  ```shell
+  watch --color --interval 5 --no-title \
+  "kubectl --namespace ${MLP_MODEL_OPS_NAMESPACE} get deployment/multimodal-embedding-model | GREP_COLORS='mt=01;92' egrep --color=always -e '^' -e '1/1     1            1'
+  echo '\nLogs(last 10 lines):'
+  kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} logs deployment/multimodal-embedding-model --tail 10"
+  ```
 
   ```
-  gcloud storage rsync gs://temporary-rag-data/flipkart_images gs://${MLP_DATA_BUCKET}/flipkart_images/
+  NAME                         READY   UP-TO-DATE   AVAILABLE   AGE
+  multimodal-embedding-model   1/1     1            1           XXXXX
   ```
 
+## Verify the model
+
 - Configure the curl job.
 
-  ```
+  ```shell
   export IMAGE_URI="$(gcloud storage ls gs://${MLP_DATA_BUCKET}/flipkart_images | head -1)"
   echo ${IMAGE_URI}
   ```
 
-  ```
+  ```shell
   git restore manifests/curl.yaml
-  sed \
-  -i -e "s|V_IMAGE_URI|${IMAGE_URI}|" \
-  manifests/curl.yaml
+  envsubst < manifests/curl.yaml | sponge manifests/curl.yaml
   ```
 
 - Create the curl job.
 
-  ```sh
+  ```shell
   kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} apply -f manifests/curl.yaml
   ```
 
 - Get the logs for the curl job.
 
+  ```shell
+  kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} logs --follow job/multimodal-curl
+  ```
+
+  The output should be similar to:
+
   ```
-  kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} logs job/multimodal-curl
+  {"multimodal_embeds":[0.5924106240272522,-0.20345857739448547,-0.14882200956344604,0.5351353883743286,0.07293718308210373,-0.15884630382061005,-0.8434014320373535,0.23294861614704132,0.1901591569185257,-0.2626086473464966,0.1809360533952713,0.6944392919540405,-0.6756398677825928,-0.4964878559112549,0.1505148708820343,0.38322848081588745,-0.16832123696804047,-1.578013300895691,-0.4120802581310272,0.7676566243171692,-0.043128449469804764,-1.2867095470428467,0.1959574967622757,0.19568923115730286,-0.3762733042240143,-0.2227010875940323,0.0216318741440773,1.1506505012512207,-0.3325354754924774,0.996836245059967,0.008820452727377415,-1.2283766269683838,1.116440773010254,0.5558998584747314,-1.233245849609375,-0.03321460634469986,-0.08853068202733994,0.36936497688293457,0.30411338806152344,-1.0682016611099243,0.685946524143219,0.46431964635849,-0.06503061205148697,-0.6995241045951843,-0.8500242829322815,-0.008173193782567978,0.04908387362957001,-0.17223168909549713,0.6433345675468445,-0.7146121263504028,-0.847214937210083,-0.46910935640335083,-0.057915765792131424,0.9733234643936157,-0.5221759676933289,0.20472489297389984,0.28791186213493347,0.5766440033912659,0.1494569033384323,0.075665183365345,-0.8721647262573242,0.3098238706588745,0.9528746604919434,-0.6286458373069763,-0.6070472598075867,-0.31975263357162476,0.5499343276023865,-0.690697193145752,-0.6751032471656799,0.017426932230591774,0.47378459572792053,1.2284945249557495,-0.10540176182985306,0.33025944232940674,0.03466665372252464,-0.2689518630504608,-0.08906934410333633,-0.39901867508888245,0.784071147441864,1.9048402309417725,0.22897273302078247,-1.060357689857483,-0.4512109160423279,0.16696316003799438,-0.2122073620557785,-0.6392346620559692,0.5564654469490051,0.2948237657546997,0.34229978919029236,-0.3559795916080475,0.17897234857082367,-0.3155965209007263,0.06322792917490005,0.6553076505661011,-0.4236254394054413,0.2814951539039612,-0.10399337857961655,-1.0849717855453491,-0.46524766087532043,0.5265820622444153,-0.039711661636829376,-0.3791617453098297,-0.8323513865470886,0.44320356845855713,-0.11838524043560028,-0.9316837191581726,0.24924667179584503,0.41757479310035706,0.038102470338344574,-1.9340052604675293,-0.025717858225107193,0.3303055167198181,1.2454990148544312,-0.1246604323387146,0.038160841912031174,-0.051868800073862076,0.8113735318183899,0.6895425319671631,0.574571430683136,0.6685034036636353,0.05573762208223343,0.9255302548408508,-0.04630526527762413,-0.6265912055969238,0.09264900535345078,1.1382992267608643,-0.12079610675573349,0.6352153420448303,-0.6007900834083557,-2.0102756023406982,-0.3941310942173004,-0.29737618565559387,0.38305985927581787,1.2179173231124878,0.3158678114414215,0.3176381289958954,-0.5689866542816162,0.415942907333374,0.6107836365699768,-0.8466510772705078,0.6475733518600464,0.2757197618484497,-1.7423759698867798,0.3180757462978363,-0.6129754185676575,0.18928393721580505,-0.28512609004974365,0.9512983560562134,0.48656705021858215,0.0262975562363863,-0.3403189182281494,0.9278243780136108,0.364020437002182,0.22098484635353088,0.9116793274879456,1.1627230644226074,0.31623464822769165,-0.8246462345123291,-0.6413137316703796,-1.249909520149231,-0.8452469110488892,-0.7048303484916687,-0.20750950276851654,-0.15846426784992218,-0.313335657119751,0.40668410062789917,1.014188528060913,0.39454370737075806,0.24124768376350403,0.08503568172454834,-0.12810391187667847,-1.1661150455474854,-0.057069167494773865,0.6312134861946106,-0.05011759698390961,0.07944561541080475,0.25088274478912354,-0.7156392931938171,0.5230376124382019,0.2992391884326935,0.5024787783622742,-0.5910110473632812,-0.6746929883956909,0.1383083611726761,1.1426094770431519,0.9035974144935608,0.2560117542743683,-0.19548475742340088,0.11429132521152496,-0.07424812763929367,0.5549275279045105,0.7408034801483154,-0.15848000347614288,-0.4925006031990051,0.3749609589576721,-0.7810171246528625,-0.8198821544647217,0.9027780294418335,-0.051281485706567764,-1.3418570756912231,-0.7416917681694031,-0.011724784038960934,0.9643922448158264,-0.9879753589630127,0.2065039575099945,1.679306983947754,-0.994052529335022,0.5123266577720642,0.6030668020248413,0.2596941292285919,-0.2549845576286316,0.25600358843803406,1.0323054790496826,-0.4192933142185211,0.006751406472176313,0.7416000366210938,0.063038170337677,0.28278666734695435,0.8160985112190247,-0.2981548309326172,0.09062497317790985,0.2493465691804886,-0.4097219705581665,1.0461561679840088,-0.48408523201942444,-0.030861463397741318,1.2159347534179688,0.4290057122707367,0.36419039964675903,-0.14557215571403503,0.2958556115627289,-1.8545191287994385,0.24612410366535187,-0.08295079320669174,-0.09110811352729797,0.604489266872406,-1.1589456796646118,0.7125701308250427,-0.02696867287158966,-0.20462752878665924,-0.3534247875213623,0.24426694214344025,-0.0363934189081192,0.29090404510498047,0.34047749638557434,-1.1260850429534912,-0.29075828194618225,-0.6621608734130859,0.16159023344516754,-0.19804151356220245,-1.2392868995666504,0.10724444687366486,0.030820529907941818,0.5301802754402161,-0.2038278430700302,-1.6347323656082153,0.4680101275444031,-0.3246447443962097,0.46105077862739563,0.2697536051273346,0.7527429461479187,1.300972580909729,-0.5727341771125793,-1.8289685249328613,2.159158229827881,-0.09365013986825943,-0.38880637288093567,0.024096481502056122,0.06783846020698547,0.8702865839004517,0.6069105267524719,-0.5236564874649048,-0.6403051018714905,-2.291926622390747,-0.9303463697433472,0.7924670577049255,0.23445598781108856,0.3862837851047516,1.5976613759994507,0.054507821798324585,-0.7957853078842163,0.037503622472286224,0.24895651638507843,0.2053094357252121,-0.3416963815689087,0.23919537663459778,0.7615337371826172,-0.3064285218715668,-0.11935160309076309,0.22598037123680115,-0.8022726774215698,0.9771227240562439,0.19433902204036713,-0.06242872029542923,-0.10598665475845337,-0.05884666368365288,-0.8490187525749207,0.31686854362487793,-0.7071172595024109,1.248034954071045,0.38836604356765747,0.09760317206382751,0.03638481721282005,-0.5906966328620911,0.5589482188224792,-0.38652920722961426,0.0038456094916909933,-0.8829982280731201,0.40402284264564514,-0.07063112407922745,0.48047932982444763,-0.505984365940094,-0.8374018669128418,-0.3932972252368927,-0.40560054779052734,-0.4289301633834839,-0.34754669666290283,-0.8173854351043701,-0.4326341152191162,-0.35555294156074524,-0.22077728807926178,-1.195460557937622,-0.12446943670511246,-0.7700096964836121,-0.05019836500287056,-0.10403405129909515,-0.8732231259346008,0.6984255909919739,0.5540904998779297,0.17234764993190765,-0.20329773426055908,-0.763291597366333,-0.7350960373878479,-0.5024373531341553,-0.07996673136949539,-0.561328113079071,0.3697052001953125,-0.7664240598678589,-0.5422042012214661,0.3661231994628906,-1.5195907354354858,0.30998358130455017,0.6927115321159363,0.9210965633392334,1.337175965309143,0.32856330275535583,-0.13148677349090576,-0.4731431007385254,1.7093837261199951,-0.9689330458641052,-0.17429831624031067,0.24648000299930573,-0.6872245073318481,0.6300744414329529,1.0549181699752808,-0.1501505821943283,0.07626151293516159,0.014565291814506054,0.6200234889984131,-0.7497552633285522,-0.13769806921482086,0.7262248992919922,-0.5686532258987427,0.3675328493118286,-0.6681644320487976,0.34052222967147827,-0.36161768436431885,0.39699244499206543,0.12775655090808868,0.8234163522720337,-0.6766327619552612,-0.37533244490623474,-0.5156548023223877,0.5635435581207275,-0.6431009769439697,-0.9025450348854065,1.0344597101211548,0.24740907549858093,-0.16893160343170166,-0.2802108824253082,-0.21651746332645416,-0.20085358619689941,-0.29250916838645935,0.9410440921783447,0.7237409949302673,0.1627369225025177,-0.2515818476676941,-0.3576256036758423,-1.1060004234313965,-0.702904999256134,-0.3310628831386566,0.713584840297699,1.4857008457183838,0.5749641060829163,0.6156712770462036,0.8447413444519043,0.2557753920555115,-0.06209149956703186,0.03618166968226433,-0.5274552702903748,0.9609156250953674,-0.5495775938034058,0.4336977005004883,-0.7704826593399048,-0.4095536470413208,-1.5273873805999756,1.0192725658416748,-0.2866295874118805,-0.18666572868824005,1.5948429107666016,-0.8420526385307312,-0.1537669450044632,-0.5174681544303894,0.5479308366775513,0.14274103939533234,0.49179983139038086,0.09275566786527634,-0.7460026144981384,0.8067554235458374,-0.19404305517673492,-1.097361445426941,1.5971705913543701,0.8954728841781616,-0.8676612973213196,-0.5234267115592957,-0.24115203320980072,0.14012376964092255,-0.9870575666427612,-0.7624554634094238,0.1638406664133072,0.12743917107582092,0.2848738729953766,0.27793166041374207,0.7528935074806213,-0.2460508793592453,-0.9296041131019592,0.24949029088020325,0.5737308859825134,0.36392930150032043,-0.4181497395038605,0.3214835524559021,-0.8477562069892883,0.5790620446205139,-0.07225629687309265,0.4959008991718292,0.20663633942604065,0.5882816314697266,0.6829842925071716,-0.03943353146314621,-1.014716386795044,0.4541794955730438,-0.793552577495575,0.7542141079902649,-0.6772245168685913,-1.0327435731887817,0.21520353853702545,0.3880807161331177,-0.44152843952178955,0.022519083693623543,0.28862690925598145,0.27966463565826416,-0.709011435508728,0.11298210918903351,-0.4660022556781769,-0.15788684785366058,0.5097206234931946,-0.4395550489425659,-0.1344972401857376,0.8646097779273987,-0.8858988881111145,-0.40364518761634827,-0.4181089401245117,-1.3044989109039307,0.6003195643424988,0.26611262559890747,0.01483498327434063,0.03671019524335861,-0.1783098578453064,-0.48309326171875,-0.021046839654445648,-0.24778850376605988,0.38099199533462524,0.46690458059310913,-0.7548868060112,0.2729513943195343,-0.22656816244125366,0.738362729549408,-0.4309113025665283,0.27661192417144775,0.3316297233104706,-0.10924235731363297,0.47210559248924255,-0.1425093412399292,-0.3397757411003113,-0.5119386315345764,-0.12244696170091629,-0.5983662605285645,0.883116602897644,0.5852846503257751,0.464468389749527,-0.5343906283378601,0.44732487201690674,0.5265068411827087,-0.8963383436203003,-0.5728183388710022,0.1321166455745697,-1.2303310632705688,-0.3244307041168213,-0.6088860034942627,-0.7326467633247375,-2.2441797256469727,-1.022038459777832,-1.1503384113311768,0.5305519104003906,0.8546370267868042,0.03200090676546097,0.6918140649795532,-0.4565896689891815,0.22333194315433502,0.4838320314884186,-0.15908588469028473,-0.42833369970321655,-1.110032558441162,-0.3613256812095642,-0.26740843057632446,-0.21148055791854858,-0.1429458111524582,2.030487537384033,-0.6748891472816467,0.32189834117889404,-0.07879403233528137,0.7610182762145996,-0.05038297548890114,-0.7954574823379517,-0.4545722007751465,-0.08980017155408859,-0.08362536132335663,-0.3386051058769226,0.7452195286750793,-0.8737925887107849,-1.2980282306671143,0.5953806042671204,0.170277401804924,0.31749382615089417,0.39990922808647156,-0.10998007655143738,-0.5869539380073547,-1.1650261878967285,1.1979150772094727,0.337417870759964,0.5308238863945007,0.2836620807647705,-0.1433805376291275,0.5892527103424072,0.6923660635948181,-1.340793490409851,0.3280276358127594,0.6116904020309448,0.7533330917358398,0.6744691133499146,0.29284119606018066,0.15764272212982178,-0.5961185693740845,-0.1882801651954651,0.42425429821014404,-1.1767736673355103,-0.0671968013048172,0.2050277590751648,0.7109279632568359,-0.12895965576171875,0.3474752604961395,-0.37441325187683105,0.055882904678583145,-0.6677221059799194,-0.24436338245868683,0.2832522988319397,-0.8741235733032227,0.4245906472206116,-0.5043145418167114,-0.008918159641325474,-0.4592393636703491,-0.14498953521251678,0.9676015377044678,-0.12521640956401825,0.05031321197748184,-0.048453543335199356,0.2883683443069458,-0.0019415427232161164,1.4985891580581665,-0.9938784241676331,-0.05295291170477867,0.8332653045654297,-0.018138568848371506,-0.20116651058197021,0.6984258890151978,0.22105801105499268,1.267309546470642,-0.3941604197025299,-0.6251749992370605,0.37921375036239624,0.1411542296409607,-0.2910098731517792,-0.8752447962760925,0.12156853079795837,-0.45161187648773193,-0.6973322033882141,-0.2819877564907074,0.26192042231559753,0.5357677936553955,-0.8252045512199402,0.2685393989086151,0.0403514988720417,-0.9001789093017578,0.5404233336448669,0.8685083985328674,-0.37962010502815247,-0.12918895483016968,-0.80665123462677,0.8366697430610657,-0.9948225617408752,0.2716467082500458,-1.4940584897994995,1.0096755027770996,0.7936453223228455,-0.37303614616394043,0.25767695903778076,0.6856227517127991,-0.9516211152076721,1.2921777963638306,0.1479228287935257,0.2804521918296814,0.7513853907585144,0.7206048965454102,-0.6532924175262451,1.1256612539291382,0.48770585656166077,-0.6284250020980835,0.18604932725429535,0.5757274627685547,0.031435687094926834,1.29680597782135,0.8489174842834473,-0.23384420573711395,-0.2413959950208664,0.4542517066001892,-0.02811950072646141,0.9807258248329163,-0.39759013056755066,0.05556461960077286,0.06736364215612411,0.4593423306941986,0.03177834302186966,0.39649659395217896,-0.5091902017593384,0.25192639231681824,-0.8313735127449036,0.4047306478023529,0.5473387837409973,-0.0973253846168518,-0.9670475721359253,-0.6689167618751526,0.4568917155265808,-0.8754829168319702,0.7289531230926514,-0.22822381556034088,0.05914885923266411,-0.7385119795799255,-2.0555343627929688,-0.20771734416484833,-0.02948600798845291,0.4359463155269623,0.5658693909645081,-0.37710848450660706,0.1891058087348938,0.5102226734161377,0.9142122268676758,-0.1439926028251648,-0.7270087003707886,0.370591938495636,0.016804920509457588,-0.8963223099708557,-0.11248458921909332,0.8208262324333191,-0.34005871415138245,0.09465346485376358,-0.29617178440093994,0.17824341356754303,0.2790500223636627,0.06746654212474823,0.21420317888259888,-0.934598445892334,-0.3456944227218628,-0.2549537420272827,0.7724777460098267,-0.8793022036552429,0.03459929674863815,0.4236677587032318,-0.2653326690196991,0.04149880260229111,-0.1393849104642868,-0.08927671611309052,0.9172636866569519,0.3412185311317444,-0.3941933214664459,0.9213281869888306,0.34835806488990784,0.005079601425677538,1.260278582572937,-0.3396773934364319,0.8979130983352661,-0.5272532105445862,0.4656248092651367,0.32662105560302734,-0.7254701256752014,0.4663284718990326,0.4497484564781189,0.021822882816195488,-1.0754454135894775,-0.1687462329864502,-0.6356937885284424,0.2550338804721832,0.3024436831474304,-0.5981853604316711,-0.18913383781909943,0.1669837385416031,-0.3048868775367737,0.831307590007782,-0.5515435338020325,0.038372695446014404,0.09054838120937347,0.10913554579019547,0.2590837776660919,-0.6824265718460083,0.2753410339355469,-0.32280510663986206,0.8028574585914612,0.3356193006038666,0.20385333895683289,-1.229520320892334,-1.5652217864990234,-0.13384152948856354,0.38042017817497253,-0.020127426832914352,-0.08777672052383423,0.6866685152053833,-0.6626774668693542,-0.13147133588790894,0.171450674533844,0.35092267394065857,0.6960228681564331,-0.26311707496643066,-0.5996111631393433,-3.0251731872558594,-0.5291815996170044,-0.5137613415718079,0.36210379004478455,-0.4394923448562622,-0.25105997920036316,-0.6869190335273743,0.23686374723911285,1.123238444328308,-1.5054799318313599,1.4507451057434082,0.049005743116140366,0.41741254925727844,0.0512530691921711,-0.3210649788379669,0.05774897336959839]}
   ```
diff --git a/use-cases/rag-pipeline/embedding-models/multimodal-embedding/manifests/curl.yaml b/use-cases/rag-pipeline/embedding-models/multimodal-embedding/manifests/curl.yaml
index 25d4d3c4..540a73d9 100644
--- a/use-cases/rag-pipeline/embedding-models/multimodal-embedding/manifests/curl.yaml
+++ b/use-cases/rag-pipeline/embedding-models/multimodal-embedding/manifests/curl.yaml
@@ -45,4 +45,4 @@ spec:
         imagePullPolicy: IfNotPresent
         env:
         - name: IMAGE_URI
-          value: V_IMAGE_URI
+          value: "${IMAGE_URI}"
diff --git a/use-cases/rag-pipeline/embedding-models/multimodal-embedding/manifests/deployment.yaml b/use-cases/rag-pipeline/embedding-models/multimodal-embedding/manifests/deployment.yaml
index cbcd6dc3..a6bd61c8 100644
--- a/use-cases/rag-pipeline/embedding-models/multimodal-embedding/manifests/deployment.yaml
+++ b/use-cases/rag-pipeline/embedding-models/multimodal-embedding/manifests/deployment.yaml
@@ -26,12 +26,12 @@ spec:
       labels:
         app: multimodal-embedding-model
     spec:
-      serviceAccountName: V_KSA
+      serviceAccountName: ${KUBERNETES_SERVICE_ACCOUNT}
       containers:
       - env:
         - name: "PORT"
           value: "5000"
-        image: V_IMAGE
+        image: ${CONTAINER_IMAGE_URL}
         imagePullPolicy: Always
         name: multimodal-embedding-model
         resources:
diff --git a/use-cases/rag-pipeline/instruction-tuned-model/README.md b/use-cases/rag-pipeline/instruction-tuned-model/README.md
index 82e2ff2f..a091ac28 100644
--- a/use-cases/rag-pipeline/instruction-tuned-model/README.md
+++ b/use-cases/rag-pipeline/instruction-tuned-model/README.md
@@ -1,4 +1,4 @@
-# Steps to deploy instruction tuned model
+# RAG: Instruction tuned model
 
 ## Prerequisites
 
@@ -9,54 +9,59 @@
 
 ## Preparation
 
-- Clone the repository
+- Clone the repository.
 
-  ```sh
+  ```shell
   git clone https://github.com/GoogleCloudPlatform/accelerated-platforms && \
   cd accelerated-platforms
   ```
 
-- Change directory to the guide directory
+- Change directory to the guide directory.
 
-  ```sh
+  ```shell
   cd use-cases/rag-pipeline/instruction-tuned-model
   ```
 
-- Ensure that your `MLP_ENVIRONMENT_FILE` is configured
+- Ensure that your `MLP_ENVIRONMENT_FILE` is configured.
 
-  ```sh
+  ```shell
   cat ${MLP_ENVIRONMENT_FILE} && \
-  source ${MLP_ENVIRONMENT_FILE}
+  set -o allexport && \
+  source ${MLP_ENVIRONMENT_FILE} && \
+  set +o allexport
   ```
 
   > You should see the various variables populated with the information specific
   > to your environment.
 
-- Get credentials for the GKE cluster
+- Get credentials for the GKE cluster.
 
-  ```sh
-  gcloud container fleet memberships get-credentials ${MLP_CLUSTER_NAME} --project ${MLP_PROJECT_ID}
+  ```shell
+  gcloud container clusters get-credentials ${MLP_CLUSTER_NAME} \
+  --dns-endpoint \
+  --project=${MLP_PROJECT_ID} \
+  --region=${MLP_REGION}
   ```
 
 ### HuggingFace access token
 
 - Set `HF_TOKEN` to your HuggingFace access token. Go to
-  https://huggingface.co/settings/tokens , click `Create new token` , provide a
+  https://huggingface.co/settings/tokens , click `Create new token`, provide a
   token name, select `Read` in token type and click `Create token`.
 
-  ```
+  ```shell
   HF_TOKEN=
   ```
 
 - Create a Kubernetes secret with your HuggingFace token.
 
-  ```sh
+  ```shell
   kubectl create secret generic hf-secret \
   --from-literal=hf_api_token=${HF_TOKEN} \
   --dry-run=client -o yaml | kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} apply  -f -
   ```
 
-## Deploy model
+## Deploy the model
 
 - Create the deployment.
 
@@ -64,24 +69,36 @@
   kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} apply -f manifests/it-model-deployment.yaml
   ```
 
-- Wait for the deployment to be ready.
+- Watch the deployment until it is ready and available.
+
+  ```shell
+  watch --color --interval 5 --no-title \
+  "kubectl --namespace ${MLP_MODEL_OPS_NAMESPACE} get deployment/rag-it-model | GREP_COLORS='mt=01;92' egrep --color=always -e '^' -e '1/1     1            1'
+  echo '\nLogs(last 10 lines):'
+  kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} logs  deployment/rag-it-model --tail 10"
+  ```
 
   ```
-  kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} wait --for=condition=ready --timeout=900s pod -l app=rag-it-model
+  NAME           READY   UP-TO-DATE   AVAILABLE   AGE
+  rag-it-model   1/1     1            1           XXXXX
   ```
 
-  When they deployment is ready you should see output similar to:
+## Verify the model
+
+- Create the curl job.
 
-  ```output
-  pod/rag-it-model-XXXXXXXXX-XXXXX condition met
+  ```shell
+  kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} apply -f manifests/curl.yaml
   ```
 
-- Verify the deployment with the curl job.
+- Get the logs for the curl job.
 
+  ```shell
+  kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} logs --follow job/it-curl
   ```
-  kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} apply -f manifests/curl.yaml
-  ```
+
+  The output should be similar to:
 
   ```
-  kubectl --namespace ${MLP_KUBERNETES_NAMESPACE} logs job/it-curl
+  {"id":"chat-XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX","object":"chat.completion","created":##########,"model":"google/gemma-2-2b-it","choices":[{"index":0,"message":{"role":"assistant","content":"You're in luck! There are tons of great cycling shorts for women out there. To help you find the perfect pair, I need a little more information about your needs and preferences. \n\n**Tell me about:**\n\n* **Your budget:** Are you looking for something affordable, mid-range, or high-end?\n* **Your riding style:** What type of cycling do you do? Road, mountain, gravel, triathlon, etc.?\n* **Your comfort priorities:** Do you prioritize chamois comfort, breathability, moisture-wicking, or something else?\n* **Your fit preferences:** Do you prefer a snug fit, a looser fit, or something in between?\n* **Your style preferences:** Do you want a specific color, pattern, or design?\n\n**Here are some popular brands known for their women's cycling shorts:**\n\n* **Specialized:** Known for their high-performance, technical shorts with excellent chamois.\n* **Pearl Izumi:** Offers a wide range of styles and features, including their popular \"Pro\" line.\n* **Castelli:** Italian brand known for their stylish and comfortable shorts.\n* **Rapha:** High-end brand with a focus on performance and style.\n* **","tool_calls":[]},"logprobs":null,"finish_reason":"length","stop_reason":null}],"usage":{"prompt_tokens":25,"total_tokens":281,"completion_tokens":256},"prompt_logprobs":null}
   ```