Deployment & Deliverybeginner11 min · 9 steps
Docker
Package the app and its exact runtime into one immutable image that runs the same on a laptop, in CI and in production.
image1,600MB
rebuild240s
requesterrordatasignal
metrics are illustrativeProblemStep 1 of 9
“It worked on my laptop”
An LLM service is a fragile stack: a specific Python version, provider SDKs, tokenizers, a PDF parser that needs system libraries, maybe CUDA. Install it by hand on three machines and you get three slightly different environments.
The code that passed tests in CI is then not quite the code that runs in production — and the difference shows up as a crash or, worse, a subtly different answer.
Watch outIf production is rebuilt from scratch, you never really tested what you shipped.
Lesson 1 of 20 in Deployment & Delivery