Deploy a Play

Deploy a FrootAI solution play to Azure with Bicep infrastructure, health checks, evaluation gates, and cost tracking.

Prerequisites

Requirement	Verify	Minimum Version
Azure CLI	`az version`	2.60+
Bicep CLI	`az bicep version`	0.25+
Azure subscription	`az account show`	Contributor role
Node.js	`node --version`	22+


az login
az account set --subscription "YOUR_SUBSCRIPTION_ID"

Step 1: Pre-Deployment Checklist


# Validate Bicep compiles
az bicep build -f solution-plays/01-enterprise-rag/infra/main.bicep
 
# Validate all primitives
npm run validate:primitives
 
# Scan for leaked secrets
bash hooks/fai-secrets-scanner/scan-secrets.sh

Step 2: Create the Resource Group


PLAY="01-enterprise-rag"
ENV="dev"
LOCATION="eastus2"
RG="rg-frootai-${PLAY}-${ENV}"
 
az group create --name "$RG" --location "$LOCATION" \
  --tags play="$PLAY" environment="$ENV" managed-by="frootai"

Step 3: Preview with What-If

⚠️

Always Preview First

Never deploy without running what-if to check for unexpected resource deletions.


az deployment group what-if \
  --resource-group "$RG" \
  --template-file "solution-plays/${PLAY}/infra/main.bicep" \
  --parameters "solution-plays/${PLAY}/infra/parameters.${ENV}.json" \
  --mode Incremental

Step 4: Deploy Infrastructure


DEPLOY_NAME="frootai-${PLAY}-$(date +%Y%m%d-%H%M%S)"
 
az deployment group create \
  --resource-group "$RG" \
  --template-file "solution-plays/${PLAY}/infra/main.bicep" \
  --parameters "solution-plays/${PLAY}/infra/parameters.${ENV}.json" \
  --name "$DEPLOY_NAME" --mode Incremental

Step 5: Deploy Application Code

For Container Apps:


ACR_NAME="acrfrootai${ENV}"
az acr build --registry "$ACR_NAME" \
  --image "frootai-${PLAY}:latest" \
  "solution-plays/${PLAY}/"

Step 6: Health Checks


APP_URL=$(az deployment group show -g "$RG" -n "$DEPLOY_NAME" \
  --query "properties.outputs.applicationUrl.value" -o tsv)
 
curl -sf "${APP_URL}/health" | jq .
# Expected: {"status":"healthy","version":"1.0.0"}

Step 7: Post-Deployment Evaluation


node engine/index.js "solution-plays/${PLAY}/fai-manifest.json" --eval

All guardrails must pass before the deployment is complete.

Cost Estimation

Resource	Dev Cost/mo	Prod Cost/mo
Azure OpenAI (GPT-4o-mini)	~$15	~$150
Azure AI Search	~$70	~$250
Container Apps	~$0-20	~$50-200
Total estimated	~$90	~$640

Rollback


# Redeploy previous version
az deployment group create --resource-group "$RG" \
  --template-file "solution-plays/${PLAY}/infra/main.bicep" \
  --parameters "solution-plays/${PLAY}/infra/parameters.${ENV}.json" \
  --name "rollback-$(date +%Y%m%d-%H%M%S)"

Troubleshooting

Symptom	Fix
`InsufficientQuota` on OpenAI	Request quota increase or switch region
Container App returns 502	Check `az containerapp logs show`
Health check timeout	Verify VNet integration and DNS