Beta

Quick Start

Deploy Ajutant into your Azure tenant in under an hour.

Overview

This guide walks you through deploying a working Ajutant instance in your Azure tenant. By the end, you’ll have:

  • A running Ajutant platform accessible via your custom domain
  • Azure AD authentication configured
  • At least one assistant ready for your team

Estimated time: 45–60 minutes.

Step 1: Register the Entra ID Application

Ajutant authenticates users through your Azure AD directory. First, create the App Registration.

# Run the prerequisite script
./scripts/register-entra-app.ps1 \
  -AppName "Ajutant" \
  -RedirectUri "https://your-domain.azurecontainerapps.io/api/auth/callback"

This script will:

  1. Create an App Registration in your Entra ID tenant
  2. Configure the required API permissions (User.Read)
  3. Generate a client secret and store it in Key Vault
  4. Output the Client ID and Tenant ID you’ll need for the next step
Save the output
Note the Client ID and Tenant ID from the script output. You’ll need these for the deployment configuration.

Step 2: Deploy Infrastructure

Ajutant’s infrastructure is defined in Bicep templates. Deploy with a single command:

az deployment group create \
  --resource-group rg-ajutant \
  --template-file infra/main.bicep \
  --parameters \
    clientId="<your-client-id>" \
    tenantId="<your-tenant-id>" \
    region="uksouth" \
    dbAdminPassword="<secure-password>"

This provisions:

  • Azure Container Apps environment
  • PostgreSQL Flexible Server with pgvector extension
  • Azure Key Vault with managed identities
  • Virtual Network with private DNS zones
  • LiteLLM proxy for model abstraction

Deployment typically takes 10–15 minutes.

Step 3: Configure Your First Model

Once infrastructure is deployed, connect at least one AI model:

  1. Open the Azure AI Foundry portal
  2. Deploy a model (we recommend starting with GPT-4o)
  3. Copy the endpoint URL and API key
  4. In the Ajutant admin panel, navigate to Settings → Models
  5. Add the model with its endpoint and key

Step 4: Create Your First Assistant

With a model connected, create an assistant:

  1. Navigate to Admin → Assistants
  2. Click Create Assistant
  3. Configure:
    • Name: e.g., “General Assistant”
    • Model: Select the model you just added
    • System prompt: Describe the assistant’s purpose and behaviour
    • Team access: Choose which Azure AD groups can access it
  4. Click Save

Your team can now access the assistant from the main dashboard.

Step 5: Verify

  1. Open Ajutant in your browser
  2. Sign in with your Azure AD credentials
  3. You should see your assistant on the dashboard
  4. Start a conversation to verify it’s working
You're live
Your private AI workspace is now running in your tenant. No data leaves your environment.

Next Steps