AI App Development from Desktop to Production

Many generative AI projects are abandoned after proof of concept — costing organizations billions. This is often due to the challenges organizations face when moving concepts to production — mismatched dependencies, untrusted packages, unvetted models, and more.

In this workshop, you’ll learn how Red Hat addresses these challenges by playing the role of a developer at Parasol Insurance.

The scenario

At Parasol, onboarding a new developer takes weeks. Getting code from a developer’s machine into production takes 3-5 days of manual handoffs. And shadow AI is a growing problem — developers are using random third-party AI services that expose the company to security and compliance risks.

Your manager has told you: "We need to move fast on AI, but we can’t keep doing it the way we have been. Prove there’s a better way."

Your mission: Build and customize an AI-powered chatbot — starting from your desktop and taking it all the way to production on OpenShift — in a single session.

What you’ll do

  • Module 1: Provisioning and exploration (~40 min) — Self-service a developer workstation from Red Hat Developer Hub, explore Podman Desktop and the AI Lab extension, run an AI recipe locally

  • Module 2: Develop and build (~40 min) — Clone and customize the Parasol Chat app, build and test locally, push to the OpenShift registry, deploy to the cluster, then push code to GitLab

  • Module 3: Deploy and verify (~40 min) — Watch the automated CI/CD pipeline build your app, observe ArgoCD deploy it, verify it’s running live

Your environment

Everything is pre-configured. Access all services through the tabs at the top of the right pane:

  • OCP Console — The Red Hat OpenShift cluster console. OpenShift is Parasol’s enterprise application platform, providing the foundation for running containers, virtual machines, and AI workloads at scale.

  • Developer Hub — Red Hat Developer Hub is Parasol’s internal developer portal, built on Backstage. It provides self-service templates, a software catalog, and a single pane of glass for CI/CD — reducing onboarding time and eliminating the need to context-switch between tools.

  • GitLab — Parasol’s source code management platform. GitLab hosts the application repositories and triggers automated pipelines when code is pushed — a key part of closing the gap between development and production.

  • ArgoCD — OpenShift GitOps powered by ArgoCD continuously monitors Git repositories and automatically syncs application deployments to the cluster, ensuring what’s in Git is what’s running in production.

Your credentials for all services:

Username:

{user}

Password:

{password}
The virtual machine desktop uses different credentials: username labuser and password labpassword.

Let’s get started!

Click the Next link below to get started with Module 1: Provisioning and exploration.