Layer 8: Cloud Deployment
Beginner explanation
Deployment is how a local prototype becomes a team-usable system. You need environments, secrets, containers, CI/CD, and a way to debug runtime failures.
Production explanation
Agentic systems add deployment complexity because they depend on model providers, vector stores, browser runtimes, background jobs, and observability infrastructure. Deployment planning should cover scaling, rollout strategy, and failure isolation.
Enterprise example
A support copilot runs as a containerized API with separate staging and production environments, managed secrets, CI checks, and dashboards for latency, errors, and spend.
Architecture diagram
TypeScript example
export function requireEnv(name: string): string {
const value = process.env[name];
if (!value) throw new Error(`Missing environment variable: ${name}`);
return value;
}
Python example
import os
def require_env(name: str) -> str:
value = os.getenv(name)
if not value:
raise RuntimeError(f"Missing environment variable: {name}")
return value
Common mistakes
- treating local
.envsetup as a deployment strategy - no staging environment for prompts, tools, and retrieval changes
- shipping browser or background dependencies without container validation
- no release rollback story
Mini exercise
Write the exact environment variables, external services, and smoke tests required to deploy one of your projects.
Project assignment
Containerize your chosen project, add a CI workflow outline, and document staging versus production configuration.
Interview questions
- What makes deploying an agentic app harder than a standard CRUD API?
- Which dependencies should be validated in CI before deployment?
- How would you control spend during an early production rollout?
Monetization angle
Deployment closes the credibility gap between demos and buying decisions. Many teams will pay for “productionization” even if they already built a prototype internally.