Tip: For best viewing in your IDE, use markdown preview (VS Code: Cmd+Shift+V on Mac, Ctrl+Shift+V on Windows/Linux)

Redis Transaction Processing Workshop

Build a real-time transaction backend with Azure Managed Redis. Store data so it can be queried in a single command and make it AI-searchable with vector embeddings.

Prerequisites

Before starting, ensure you have the required software installed:

• Prerequisites Guide — All required software and installation instructions

What You'll Learn

• Redis Streams — Ingest transactions in real-time
• Redis Lists — Retrieve recent transactions in order
• Redis JSON — Store and query transaction details
• Sorted Sets — Rank spending by category and merchant
• TimeSeries — Track spending trends over time
• Vector Search — Search transactions by meaning, not keywords

Get Started

Option A: GitHub Codespaces (Recommended)

The fastest way to start — no local installation required!

1. Click the button above (or go to Code → Codespaces → Create)
2. Wait ~2 minutes for the container to build
3. Login to Azure:

   azd auth login
   az login --use-device-code

4. Deploy:

   azd up

   Note: If deployment fails with ParentResourceNotFound, run azd provision && azd deploy to complete. This is a known ARM race condition.

Everything is pre-installed: azd, az, python, pwsh.

---

Option B: Local Development

Step 1: Login to Azure

azd auth login
az login

Step 2: Deploy to Azure

azd up

You'll be prompted for:

Prompt	What to Enter	Example
Environment name	A unique name for your deployment (lowercase, no spaces)	john-workshop
Azure subscription	Select from your available subscriptions (use arrow keys)	My Subscription

Note: Resources deploy to westus3 by default (best Azure Managed Redis availability). To use a different region: azd env set AZURE_LOCATION eastus && azd up

⚠️ Known Issue: ARM Race Condition

Azure Managed Redis v2 API may report ParentResourceNotFound during first deployment even though resources are created successfully. This is an ARM timing issue.

If deployment fails:

• Run azd provision && azd deploy to complete
• The resources are usually already deployed — the error is cosmetic

Typical deployment time: 15-25 minutes (first run may fail and require a second azd provision)

Tip: The environment name is used for:

• Resource group: rg-<name> (e.g., rg-john-workshop)
• azd environment reference (for managing multiple deployments)

Other resources get auto-generated names with a unique hash (e.g., redis-abc123xyz).

First deployment takes 15-25 minutes. Grab a coffee! ☕

Step 3: Access Your Application

After deployment, you'll see URLs like:

UI URL:         https://ui.xxx.azurecontainerapps.io
API URL:        https://api.xxx.azurecontainerapps.io
Redis Insight:  https://redis-insight.xxx.azurecontainerapps.io

Open the UI URL to start the workshop!

---

Deploy Your Code Changes

After editing any module file, deploy in ~10 seconds:

./sync-and-restart.sh

This uploads your processor/modules/*.py to Azure and restarts the processor.

---

Workshop Modules

Module	Topic	Type
Module 0	Explore Your Deployment	Portal Walkthrough
Module 1-5	Build the Transaction Processor	Hands-on Coding
Module 6	Observability with AMR	Portal Walkthrough

Start with Module 0 to understand what was deployed, then head to the coding modules.

---

Clean Up

When done, delete all Azure resources:

azd down

📚 More Documentation:

• Prerequisites — Required software
• Azure Deployment — Full deployment guide & troubleshooting