Deploying a Dash ML App on AWS

The problem

Jacqueline McShan had a working Dash Plotly ML application on her machine and wanted it on the public web under her own domain. Getting from a local script to a hosted service meant settling the dependency versions, documenting how the app is installed and run, and choosing AWS services that would serve it reliably without a heavy setup.

What I built

• A revised set of documentation describing how to install, configure and run the application on a server.
• Pinned package versions, so the environment on AWS matches the one the app was developed in and can be rebuilt at any time.
• The code edits the application needed to run as a hosted service rather than a local development server.
• An AWS EC2 instance hosting the Dash application.
• Route 53 DNS configuration that redirects requests for the custom domain to the instance.

How it was delivered

I treated the deployment as a small productionisation job rather than a copy-and-run. Reproducible dependencies and clear documentation came first, because an app that only runs on one laptop cannot be rehosted or fixed later. Once the package set was stable, the EC2 instance and the Route 53 records were straightforward.

Outcome

The application is hosted on EC2 and reachable at jacquelinemcshan.net through Route 53, with documentation and pinned versions that make the hosting reproducible. The client has a setup that can be rebuilt from the docs rather than one that only works on a single machine.