Deployments

Local demos ≠ production. Deployment covers env vars, secrets, health checks, rollbacks, and zero-downtime updates.

Deployment is moving from cooking at home to opening a restaurant — same recipe, but health inspections, staff, and supply chains matter.

Scenario

Merge to main triggers GitHub Actions: test → build image → deploy to staging → run eval suite → promote to prod.

Solution

In Production Agent Engineering, apply Deployments to this scenario: Merge to main triggers GitHub Actions: test → build image → deploy to staging → run eval suite → promote to prod. Identify the inputs, run the technique, validate the output, and note one thing you would monitor in production.

Do this before moving to the next module — reading alone is not enough.

Open the Code Walkthrough below and run it locally. Change one parameter related to Deployments (e.g. model, temperature, top_k, or tool name), observe the difference in output, and write 2–3 sentences explaining what changed.

Highlighted lines show where Deployments happens in the code.

1# Deployments — minimal example2from openai import OpenAI3
4client = OpenAI()  # create API client5
6# Ask the model to explain this topic7response = client.chat.completions.create(  # core API call for Deployments8    model="gpt-4o-mini",9    messages=[10        {"role": "system", "content": "You explain deployments clearly."},11        {"role": "user", "content": f"What is deployments?"},12    ],13    temperature=0,14)15print(response.choices[0].message.content)  # show output for debugging

• Treating Deployments as a black box without evaluation
• Ignoring cost and latency in production
• Skipping error handling for deployments

Quick recap — the most important points from this module.