Cloud, deployment & production

Deploy an AI service to the cloud

Ship an LLM/ML API on Cloud Run / AWS Lambda-ECS / Azure with HTTPS, env config and logs.

~12 focused hours·intermediate

Tools: Cloud Run, AWS Lambda/ECS, Azure Container Apps, env config/secrets manager, HTTPS/TLS

What employers mean

You should be able to…

  1. Take a containerized API and deploy it to Cloud Run/Lambda/ECS with a public HTTPS URL
  2. Configure environment variables and secrets for a deployed service (not hardcoded)
  3. Set up autoscaling limits so a traffic spike doesn't blow the cloud bill
  4. Read and act on deployed service logs to debug a production error
  5. Roll back a bad deploy to the previous working version
  6. Set up a health-check endpoint the platform uses to know the service is alive
  7. Choose between serverless and container-based deploy for a given latency/cost profile

Needs first: Containerize an application with Docker

Learn — free, link-checked

The few resources that matter

Read · beginner · 40 min · docs.docker.com

Containerize a Python application

A Python-specific walkthrough for the exact Dockerfile patterns (deps, layers, entrypoint) you'll use to containerize an AI service. — Docker
Course · beginner · 90 min · cloudskillsboost.google

Google Cloud Skills Boost — Learning Paths

Free guided, hands-on labs (not just docs) for GCP and Vertex AI, with quizzes and a shareable badge for your portfolio. — Google Cloud
Read · intermediate · 30 min · cloud.google.com

Deploy a Python service to Cloud Run

The fastest free path from a Python API to a public HTTPS endpoint with autoscaling — the most common 'ship it' step in Indian AI job tests. — Google Cloud
Read · intermediate · 30 min · docs.aws.amazon.com

Building Lambda functions with Python

Covers the AWS alternative to Cloud Run — packaging, handlers and env config for a serverless Python deployment. — AWS
Read · intermediate · 30 min · learn.microsoft.com

Deploy your first container app

The Azure equivalent deploy path — useful since several Indian enterprise/BFSI employers standardize on Azure. — Microsoft Learn
Read · intermediate · 30 min · 12factor.net

The Twelve-Factor App

The config/state/logging principles behind every scalable cloud service, including the ones you'll deploy for AI inference. — Heroku / community
Practice

Deploy a Public AI Endpoint with Logs and Rollback

Deploy the containerized ticket/RAG API to Cloud Run (or Lambda/Azure Container Apps), with env-based secrets, a /health endpoint, autoscaling min/max set, and structured logging. Simulate a bad deploy and practice rolling back to the previous revision, documenting the exact steps.

Done when
  • Service is reachable over public HTTPS and returns correct responses
  • Secrets are injected via the platform's secret manager/env config, not the image
  • /health endpoint returns 200 when dependencies are up and a non-200 when a dependency is down
  • A documented rollback: deploy a broken revision on purpose, then roll back, with before/after logs
Prove it

Evidence a recruiter can check

  • Public HTTPS URL of the deployed service (or a screen recording if later torn down for cost)
  • Deployment config (cloudbuild.yaml/GitHub Action) in a public repo
  • Logs/screenshot showing the rollback from a broken revision to a working one
Interview

Questions you'll get asked

  1. Walk me through deploying a FastAPI service to Cloud Run from a Dockerfile.
  2. How do you manage secrets for a production deployment without putting them in code?
  3. A deployed service is returning 500s intermittently — how do you find out why?
  4. What's the difference between deploying to Lambda vs ECS vs a VM, and when would you pick each?
  5. How would you set up autoscaling so a viral traffic spike doesn't 10x your cloud bill?
  6. How do you roll back a bad deployment in under 5 minutes?
  7. What does your health-check endpoint verify, and why does that matter for uptime?
See where you stand for Machine Learning Engineer