Hands-on lesson

Make AI Return Structured Data

Request JSON from the model, parse it safely, validate required fields, and handle malformed output.

12 minFree Core9 Hands-On AI Lessonsbeginner

What you'll build

A structured extraction script

Validated name, role, and skills fields

A controlled malformed-JSON failure

Model output

format=json

→

JSON parser

json.loads(raw)

→

Validation

check required fields

→

Valid fields

name, role, skills

--broken flag

raw = '{broken'

→

json.loads fails

JSONDecodeError

→

except block

except JSONDecodeError

→

Malformed state

Invalid model JSON

Check

Before You Build

2 quick questions. This is diagnostic only — it does not block the lesson.

1 / 4Core

Why use json.loads?

json.loads converts JSON text into Python objects.

2 / 4Core

Why validate fields after parsing?

Syntactically valid JSON may still omit required data.

understand

build

break

debug

improve

deploy

Mission

Request JSON from the model, parse it safely, validate required fields, and handle malformed output.

Understand

Text becomes application data only after parsing and validation. A model request is untrusted input: require JSON, check its shape, and fail clearly when it is malformed.

Build

What you're doing

Create structured.py, save it, then request and validate JSON.

Why it matters

Parsing plus explicit field checks turns model text into usable application data.

Create, open, paste, and save~/enjyra-free-ai/app/structured.py

Where to work VS Code FileLanguage python

Create and open the fileRun this from ~/enjyra-free-ai.

mkdir -p app
touch app/structured.py
code app/structured.py

mkdir -p app
touch app/structured.py
code app/structured.py

New-Item -ItemType Directory -Force -Path "app" | Out-Null
if (-not (Test-Path "app\structured.py")) { New-Item -ItemType File -Path "app\structured.py" | Out-Null }
code "app\structured.py"

Before You Run

Your terminal is open in ~/enjyra-free-ai.
Python 3 is installed; activate .venv first when the lesson has already created it.
Visual Studio Code is installed; the optional code shell command is available, or you can open the file manually.

Why You're Running This

app/structured.py is the project’s python file at ~/enjyra-free-ai/app/structured.py. It must exist before you can paste the supplied content. The commands create its folder when needed, create the file, and open it in VS Code; the next stage tells you what to paste and reminds you to save it.

Command Breakdown

mkdir -p / New-Item Directory: Creates the requested folder and any missing parent folders without failing when it already exists.
touch / New-Item File: Creates an empty file at the requested path without requiring code to exist first.
code: Opens the file or folder in Visual Studio Code from the terminal.

Expected Result

app/
└── structured.py

app/structured.py exists and is ready for the supplied content.

What Changed

Beforeapp/structured.py may not exist yet, or may still contain its previous content.

Afterapp/structured.py now exists in the lesson workspace and is ready for the next instruction.

Verify It

macOSls -l app/structured.py

Linuxls -l app/structured.py

Windows PowerShellGet-Item "app\structured.py"

Confirm the file exists at the expected path.

If It Fails

Check that the terminal is in the lesson project, the path spelling is exact, and VS Code is installed. If the code command is unavailable, use File → Open Folder in VS Code.

If code is unavailable, open VS Code, choose File → Open Folder, select enjyra-free-ai, then create app/structured.py in the Explorer.

Paste this contentCopy the complete block into app/structured.py.

import json
import sys
import requests

text = "Arun is a DevOps Engineer skilled in Docker, Kubernetes, and Azure."
prompt = f"Return only JSON with name, role, and skills for: {text}"
response = requests.post("http://localhost:11434/api/generate", json={"model":"llama3.2:3b","prompt":prompt,"format":"json","stream":False}, timeout=120)
response.raise_for_status()
raw = response.json()["response"] if "--broken" not in sys.argv else "{broken"
try:
    data = json.loads(raw)
except json.JSONDecodeError as exc:
    raise SystemExit(f"Invalid model JSON: {exc.msg}") from exc
required = {"name", "role", "skills"}
missing = required - data.keys()
if missing or not isinstance(data["skills"], list):
    raise SystemExit(f"Invalid schema; missing={sorted(missing)}")
print(json.dumps(data, indent=2))

Save before continuing: press Cmd + S on macOS or Ctrl + S on Linux and Windows.

Verify the saved fileThe command should print the content you just saved.

cat app/structured.py

cat app/structured.py

Get-Content "app\structured.py"

Before You Run

Your terminal is open in ~/enjyra-free-ai.
Python 3 is installed; activate .venv first when the lesson has already created it.

Why You're Running This

Reading app/structured.py back from disk proves that the python content was saved at the correct path, rather than remaining only in an unsaved editor tab.

Command Breakdown

cat / Get-Content: Reads the saved file and prints its contents so you can verify it.

Expected Result

The terminal prints the complete content you pasted into app/structured.py.

What Changed

BeforeThe current state has not yet been checked.

AfterNo persistent state changed; the command printed evidence you can compare with the expected result.

Verify It

macOScat app/structured.py

Linuxcat app/structured.py

Windows PowerShellGet-Content "app\structured.py"

Run the command and compare its output with the Expected Result block.

If It Fails

Save the file, confirm the filename and capitalisation, and check the current folder with pwd on macOS/Linux or Get-Location on Windows.

If it fails

Check that your terminal is in ~/enjyra-free-ai, the filename and capitalisation match exactly, and the file was saved. Use pwd on macOS/Linux or Get-Location on Windows to check your current folder.

TerminalChoose macOS, Linux, Windows, or Explain

Where to run Terminal or VS Code TerminalWorking directory ~/enjyra-free-ai

What to run

cd ~/enjyra-free-ai && . .venv/bin/activate && python3 app/structured.py

cd ~/enjyra-free-ai && . .venv/bin/activate && python3 app/structured.py

Set-Location $HOME\enjyra-free-ai
.\.venv\Scripts\Activate.ps1
py app\structured.py

Before You Run

Your terminal is open in ~/enjyra-free-ai.
Python 3 is installed; activate .venv first when the lesson has already created it.

Why You're Running This

Running the program proves that the saved code and its local dependencies work together.

Command Breakdown

cd / Set-Location: Changes the terminal’s current working directory before the next command runs.
.: Runs the . tool with the shown arguments to complete this lesson step.
python: Runs the selected Python module or script with the supplied arguments.
activate: Makes the project’s virtual environment the active Python environment for this terminal session.
&&: Runs the next command only when the previous command succeeds.

Expected Result

{
  "name": "Arun",
  "role": "DevOps Engineer",
  "skills": [...]
}

The exact wording may vary, but valid JSON contains all three required fields.

What Changed

BeforeThe saved program has not yet been executed for this verification step.

AfterThe program ran and produced output or started the local process described by the lesson.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

No such file or directory → verify the current folder and the path spelling.
Permission denied → work inside your own lesson folder and confirm it is writable.

What you should see

{
  "name": "Arun",
  "role": "DevOps Engineer",
  "skills": [...]
}

Success looks like: The exact wording may vary, but valid JSON contains all three required fields.

If it fails

command not found → verify the required tool is installed and reopen the terminal.
No such file or directory → check your current folder with pwd or Get-Location.
Permission denied → confirm the lesson workspace is writable.

What changed

app/structured.py exists and prints validated JSON.

Break

TerminalChoose macOS, Linux, Windows, or Explain

Where to run Terminal or VS Code TerminalWorking directory ~/enjyra-free-ai

What to run

cd ~/enjyra-free-ai && . .venv/bin/activate && python3 app/structured.py --broken

cd ~/enjyra-free-ai && . .venv/bin/activate && python3 app/structured.py --broken

Set-Location $HOME\enjyra-free-ai
.\.venv\Scripts\Activate.ps1
py app\structured.py --broken

Before You Run

Your terminal is open in ~/enjyra-free-ai.
Python 3 is installed; activate .venv first when the lesson has already created it.

Why You're Running This

Running the program proves that the saved code and its local dependencies work together.

Command Breakdown

cd / Set-Location: Changes the terminal’s current working directory before the next command runs.
.: Runs the . tool with the shown arguments to complete this lesson step.
python: Runs the selected Python module or script with the supplied arguments.
activate: Makes the project’s virtual environment the active Python environment for this terminal session.
&&: Runs the next command only when the previous command succeeds.

Expected Result

Invalid model JSON

The deliberate malformed response is caught.

What Changed

BeforeThe saved program has not yet been executed for this verification step.

AfterThe program ran and produced output or started the local process described by the lesson.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

No such file or directory → verify the current folder and the path spelling.
Permission denied → work inside your own lesson folder and confirm it is writable.

What you should see

Invalid model JSON

Success looks like: The deliberate malformed response is caught.

If it fails

Debug

TerminalChoose macOS, Linux, Windows, or Explain

Where to run Terminal or VS Code TerminalWorking directory ~/enjyra-free-ai

What to run

cd ~/enjyra-free-ai && . .venv/bin/activate && python3 -m json.tool <<<'{"name":"Arun","role":"DevOps Engineer","skills":[]}'

cd ~/enjyra-free-ai && . .venv/bin/activate && python3 -m json.tool <<<'{"name":"Arun","role":"DevOps Engineer","skills":[]}'

Set-Location $HOME\enjyra-free-ai
.\.venv\Scripts\Activate.ps1
'{"name":"Arun","role":"DevOps Engineer","skills":[]}' | py -m json.tool

Before You Run

Your terminal is open in ~/enjyra-free-ai.
Python 3 is installed; activate .venv first when the lesson has already created it.

Why You're Running This

Running the program proves that the saved code and its local dependencies work together.

Command Breakdown

cd / Set-Location: Changes the terminal’s current working directory before the next command runs.
.: Runs the . tool with the shown arguments to complete this lesson step.
python: Runs the selected Python module or script with the supplied arguments.
activate: Makes the project’s virtual environment the active Python environment for this terminal session.
'{"name":"Arun","role":"DevOps: Runs the '{"name":"Arun","role":"DevOps tool with the shown arguments to complete this lesson step.
&&: Runs the next command only when the previous command succeeds.

Expected Result

Formatted valid JSON

The parser distinguishes valid JSON from malformed text.

What Changed

BeforeThe saved program has not yet been executed for this verification step.

AfterThe program ran and produced output or started the local process described by the lesson.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

No such file or directory → verify the current folder and the path spelling.
Permission denied → work inside your own lesson folder and confirm it is writable.

What you should see

Formatted valid JSON

Success looks like: The parser distinguishes valid JSON from malformed text.

If it fails

Improve

Production systems should use a formal schema library, but explicit required-field and type checks teach the essential boundary first.

Verify

TerminalChoose macOS, Linux, Windows, or Explain

Where to run Terminal or VS Code TerminalWorking directory ~/enjyra-free-ai

What to run

cd ~/enjyra-free-ai && . .venv/bin/activate && python3 app/structured.py

cd ~/enjyra-free-ai && . .venv/bin/activate && python3 app/structured.py

Set-Location $HOME\enjyra-free-ai
.\.venv\Scripts\Activate.ps1
py app\structured.py

Before You Run

Your terminal is open in ~/enjyra-free-ai.
Python 3 is installed; activate .venv first when the lesson has already created it.

Why You're Running This

Running the program proves that the saved code and its local dependencies work together.

Command Breakdown

cd / Set-Location: Changes the terminal’s current working directory before the next command runs.
.: Runs the . tool with the shown arguments to complete this lesson step.
python: Runs the selected Python module or script with the supplied arguments.
activate: Makes the project’s virtual environment the active Python environment for this terminal session.
&&: Runs the next command only when the previous command succeeds.

Expected Result

<validated JSON>

The normal model path parses and validates again.

What Changed

BeforeThe saved program has not yet been executed for this verification step.

AfterThe program ran and produced output or started the local process described by the lesson.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

No such file or directory → verify the current folder and the path spelling.
Permission denied → work inside your own lesson folder and confirm it is writable.

What you should see

<validated JSON>

Success looks like: The normal model path parses and validates again.

If it fails

Challenge

Repeat the verification from a fresh terminal and explain which observable signal proves each dependency is healthy.

Ship

A model response converted into validated Python application data.

Next Step

Continue to Lesson 06: Give AI Your Own Documents.

03 · Prerequisites

Before You Start the Lab

Confirm your machine has everything required for this lesson. Continue only when the verification commands succeed.

This lesson needs

DockerPython 3VS Code — requiredOllamaInternet: only for first-time pullsGPU: optional

ENJYRA Lab Setup completed

Prior lesson 04 completed

Docker Desktop or Docker Engine running

Terminal access

Python 3 installed

Ollama runtime available

llama3.2:3b model available

Visual Studio Code — required for this guided lab

Disk10 GB free if the runtime or model is not cached

Memory8 GB recommended

GPUOptional

InternetRequired only for uncached dependencies or images

Docker setup

Docker CLI

Docker Engine

Test container

Ollama API

Docker not installed? Show setup instructions ▾

Check your Mac's chip architecture (Apple Silicon vs Intel) before downloading.
Where to runTerminalPlatformmacOS
COMMAND
```
uname -m
```
Expected output"arm64" (Apple Silicon) or "x86_64" (Intel)
If error
If nothing prints, open Terminal from Applications → Utilities and try again.
Download Docker Desktop for your chip from docs.docker.com/desktop/setup/install/mac-install and open the downloaded .dmg, then drag Docker to Applications.
Launch Docker Desktop from Applications and wait for the whale icon in the menu bar to show "Docker Desktop is running".
Verify the CLI is installed.
Where to runTerminalPlatformmacOS
COMMAND
```
docker --version
```
Expected outputDocker version 2x.x.x, build xxxxxxx
If error
If “command not found”, reopen Terminal after installing so your PATH refreshes, or reinstall Docker Desktop.
Verify the engine is actually running (not just installed).
Where to runTerminalPlatformmacOS
COMMAND
```
docker info
```
Expected outputA block of server information, including Server Version and Containers
If error
If it says “Cannot connect to the Docker daemon”, open Docker Desktop and wait for the whale icon to stop animating.
Run a real test container end-to-end.
Where to runTerminalPlatformmacOS
COMMAND
```
docker run --rm hello-world
```
Expected output"Hello from Docker!" followed by an explanation of what happened
If error
If the image can't be pulled, check your internet connection; this is the only step that needs it.

Download Docker Desktop for Windows from docs.docker.com/desktop/setup/install/windows-install and run the installer. Docker Desktop will guide you through any compatible Windows backend requirement; this lesson does not require a WSL terminal.
If the installer asks you to enable a Windows feature or restart, follow that prompt and finish the Docker Desktop setup before continuing.
Start Docker Desktop from the Start menu and wait for the whale icon in the system tray to report "Docker Desktop is running".
Verify the CLI, using PowerShell or the built-in VS Code terminal (not a Bash-syntax terminal).
Where to runPowerShell or VS Code terminalPlatformWindows
COMMAND
```
docker --version
```
Expected outputDocker version 2x.x.x, build xxxxxxx
If error
If “not recognized”, close and reopen your terminal so PATH changes take effect.
Verify the engine is running.
Where to runPowerShell or VS Code terminalPlatformWindows
COMMAND
```
docker info
```
Expected outputA block of server information, including Server Version and Containers
If error
If it can't connect to the daemon, open Docker Desktop, wait until it reports Running, and review Docker Desktop's own Troubleshoot panel.
Run a real test container.
Where to runPowerShell or VS Code terminalPlatformWindows
COMMAND
```
docker run --rm hello-world
```
Expected output"Hello from Docker!" followed by an explanation of what happened
If error
If the pull hangs, check your internet connection and that Docker Desktop shows as running, not starting.

Prefer Docker Engine (not Docker Desktop) on Linux. Follow the official instructions for your distribution at docs.docker.com/engine/install — for Ubuntu specifically, use docs.docker.com/engine/install/ubuntu.
Start and enable the Docker service.
Where to runTerminalPlatformLinux
COMMAND
```
sudo systemctl enable --now docker
```
Expected outputNo output on success
If error
If systemctl isn't available, your distro may use a different init system — check its Docker Engine install page.
Confirm the service is active.
Where to runTerminalPlatformLinux
COMMAND
```
systemctl status docker
```
Expected outputActive: active (running)
If error
If it shows “inactive” or “failed”, re-run the enable command above and check the install steps were completed.
Verify the CLI.
Where to runTerminalPlatformLinux
COMMAND
```
docker --version
```
Expected outputDocker version 2x.x.x, build xxxxxxx
If error
If “command not found”, the Engine install likely didn't finish — re-check the distro-specific install page.
Verify the engine responds.
Where to runTerminalPlatformLinux
COMMAND
```
docker info
```
Expected outputA block of server information, including Server Version and Containers
If error
A permission-denied error here usually means your user isn't in the docker group yet — see the note below rather than running commands as root.
Run a real test container.
Where to runTerminalPlatformLinux
COMMAND
```
docker run --rm hello-world
```
Expected output"Hello from Docker!" followed by an explanation of what happened
If error
If you see a permission error, add your user to the docker group yourself (sudo usermod -aG docker $USER, then log out and back in) — we won't do this for you automatically.

Note: running Docker commands as root or with a global chmod on the socket is not recommended. Add your user to the docker group instead, as shown above.

Before You Run

Install Docker Desktop on macOS or Windows, or Docker Engine/Desktop using the official instructions for your Linux distribution. Start Docker and open Terminal or native Windows PowerShell.

Why Do We Need Docker?

The ENJYRA AWS, Azure, and GCP labs run locally inside Docker containers. No real AWS, Azure, or GCP account is required.

Command Breakdown

docker --version: Confirms the Docker command-line client is installed.
docker info: Asks the Docker Engine for server information, proving it is running and reachable.
docker run --rm hello-world: Runs a small official test container end to end and removes it when finished.

Expected Result

Docker version 2x.x.x
docker info: server details
Hello from Docker!

What Changed

BeforeDocker is not yet confirmed.

AfterDocker CLI and Docker Engine are available to launch the local lab.

Verify It

macOSdocker --version docker info

Linuxdocker --version docker info

Windows PowerShelldocker --version docker info

If It Fails

Open Docker Desktop or start Docker Engine, wait until it reports running, and retry.
Open a new terminal after installation so PATH changes are loaded.
On Linux, follow the official instructions for your distribution rather than assuming Ubuntu commands apply.

Lesson-specific check

Where to runTerminalPlatformmacOS / Windows / Linux

COMMAND

curl http://localhost:11434/api/version

Expected output{"version":"0.x.x"}

If error

If connection is refused, the Ollama container isn't running yet — start it with this lesson's start command, then retry.

Verify your setup

Run each command in your own terminal. This page cannot run commands on your machine, so tick a box only after you see the expected result.

Docker CLIConfirms the Docker command-line client is installed.

COMMAND

docker --version

docker --version

docker --version

Before You Run

Your terminal is open in Any terminal folder.
Docker Desktop or Docker Engine is running.

Why You're Running This

Confirms the Docker command-line client is installed.

Command Breakdown

docker: Runs the docker tool with the shown arguments to complete this lesson step.

Expected Result

A Docker version is printed

What Changed

BeforeThe action shown by this card has not yet been completed or verified.

AfterThe command has completed and the next lesson step has the state it needs.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

Confirm Docker CLI is installed or running, then retry the command in a new terminal.

Expected resultA Docker version is printed

I got the expected result

Docker EngineConfirms Docker is actually running, not only installed.

COMMAND

docker info

docker info

docker info

Before You Run

Your terminal is open in Any terminal folder.
Docker Desktop or Docker Engine is running.

Why You're Running This

Confirms Docker is actually running, not only installed.

Command Breakdown

docker: Runs the docker tool with the shown arguments to complete this lesson step.

Expected Result

Docker server information is returned

What Changed

BeforeThe action shown by this card has not yet been completed or verified.

AfterThe command has completed and the next lesson step has the state it needs.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

Confirm Docker Engine is installed or running, then retry the command in a new terminal.

Expected resultDocker server information is returned

I got the expected result

Ollama APIConfirms the local Ollama runtime is reachable on its API port.

COMMAND

curl http://localhost:11434/api/version

curl http://localhost:11434/api/version

curl http://localhost:11434/api/version

Before You Run

Your terminal is open in Any terminal folder.
The local service named in the URL is already running.

Why You're Running This

Confirms the local Ollama runtime is reachable on its API port.

Command Breakdown

curl: Makes an HTTP request to the local service and prints its response.

Expected Result

The Ollama version endpoint responds

What Changed

BeforeThe action shown by this card has not yet been completed or verified.

AfterThe command has completed and the next lesson step has the state it needs.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

Confirm Ollama API is installed or running, then retry the command in a new terminal.

Expected resultThe Ollama version endpoint responds

I got the expected result

PythonConfirms Python 3 is installed. On Windows PowerShell use py --version or python --version if python3 is not found.

COMMAND

python3 --version

python3 --version

py --version
# If py is unavailable:
python --version

Before You Run

Your terminal is open in Any terminal folder.
Python 3 is installed; activate .venv first when the lesson has already created it.

Why You're Running This

Confirms Python 3 is installed. On Windows PowerShell use py --version or python --version if python3 is not found.

Command Breakdown

python: Runs the selected Python module or script with the supplied arguments.
py: Runs the py tool with the shown arguments to complete this lesson step.
#: Runs the # tool with the shown arguments to complete this lesson step.

Expected Result

A supported Python 3.x version is printed

What Changed

BeforeThe saved program has not yet been executed for this verification step.

AfterThe program ran and produced output or started the local process described by the lesson.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

Confirm Python is installed or running, then retry the command in a new terminal.

Expected resultA supported Python 3.x version is printed

I got the expected result

Visual Studio CodeOpen Visual Studio Code from your applications. The optional code --version shell command is not required; VS Code works without it.

Expected resultVisual Studio Code opens successfully

Visual Studio Code ready

Ready to start?

Docker CLI verified
Docker Engine running
Ollama API responds
Python 3 verified
Visual Studio Code ready

Need help with a prerequisite?

Docker not running? Open Docker Desktop or start Docker Engine, then run docker info.

Permission denied? Confirm your account can access the Docker daemon; do not use a global prune.

Port already in use? Use docker ps to identify the container that owns the port.

Ollama API not responding? Run docker ps, then restart the Ollama service from the lesson command.

Model not found? Run docker exec enjyra-ollama ollama list.

04 · ENJYRA Lab

Launch the ENJYRA Lab

Start only after the prerequisite checks above are complete.

This button confirms readiness; commands still run in your own terminal.

ENJYRA Lab · 05-structured-output-v6.0.0

AvailabilityREMOTE VERIFIED

RuntimeDocker + Ollama

Imageenjyra/free-python-labs:05-structured-output-v6.0.0

PlatformsDocker Desktop · Docker Engine

Architecturelinux/arm64

Start lab

Where to runTerminalWorking directoryAny terminal folderPlatformmacOS · Linux · Windows

COMMAND

docker run --rm --name enjyra-v6-structured-output-05 enjyra/free-python-labs:05-structured-output-v6.0.0

docker run --rm --name enjyra-v6-structured-output-05 enjyra/free-python-labs:05-structured-output-v6.0.0

docker run --rm --name enjyra-v6-structured-output-05 enjyra/free-python-labs:05-structured-output-v6.0.0

Before You Run

Your terminal is open in Any terminal folder.
Docker Desktop or Docker Engine is running.
Python 3 is installed; activate .venv first when the lesson has already created it.

Why You're Running This

Creates and starts the verified ENJYRA lesson container on your machine.

Command Breakdown

docker run: Creates and starts a container from the named image using the supplied ports, environment values, and runtime options.
--rm: Automatically removes the temporary container after it stops.
--name enjyra-v6-structured-output-05: Assigns a predictable lesson-scoped name so the container can be inspected or removed safely.
enjyra/free-python-labs:05-structured-output-v6.0.0: Names the verified ENJYRA lesson image and its pinned lesson version.

Expected Result

The lesson container starts successfully.

What Changed

BeforeThe requested local service or lab containers are not yet confirmed running.

AfterThe local service or lab containers are running in the background and can be verified.

Verify It

macOSdocker ps

Linuxdocker ps

Windows PowerShelldocker ps

Confirm the expected container is listed as running or healthy.

If It Fails

Check the current folder, confirm Docker is running when required, and verify the image or script name before retrying.

What this command does Creates and starts the verified ENJYRA lesson container on your machine.

Expected result

The lesson container starts successfully.

If it fails

Check the current folder, confirm Docker is running when required, and verify the image or script name before retrying.

Verify running state

Where to runTerminalWorking directoryAny terminal folderPlatformmacOS · Linux · Windows

COMMAND

docker image inspect enjyra/free-python-labs:05-structured-output-v6.0.0 --format '{{.Architecture}}'

docker image inspect enjyra/free-python-labs:05-structured-output-v6.0.0 --format '{{.Architecture}}'

docker image inspect enjyra/free-python-labs:05-structured-output-v6.0.0 --format '{{.Architecture}}'

Before You Run

Your terminal is open in Any terminal folder.
Docker Desktop or Docker Engine is running.
Python 3 is installed; activate .venv first when the lesson has already created it.

Why You're Running This

Reads the local image metadata and prints only its architecture so you can confirm that the expected ENJYRA image exists.

Command Breakdown

docker image inspect: Reads metadata for the named local image without creating or starting a container.
enjyra/free-python-labs:05-structured-output-v6.0.0: Identifies the exact ENJYRA image and lesson tag to inspect.
--format: Prints only the requested metadata field instead of the complete image record.

Expected Result

amd64 or arm64

What Changed

BeforeThe current state has not yet been checked.

AfterNo persistent state changed; the command printed evidence you can compare with the expected result.

Verify It

macOSdocker image inspect enjyra/free-python-labs:05-structured-output-v6.0.0 --format '{{.Architecture}}'

Linuxdocker image inspect enjyra/free-python-labs:05-structured-output-v6.0.0 --format '{{.Architecture}}'

Windows PowerShelldocker image inspect enjyra/free-python-labs:05-structured-output-v6.0.0 --format '{{.Architecture}}'

Run the command and compare its output with the Expected Result block.

If It Fails

Confirm the start command completed first, then compare the image, container, port, or script name exactly with this lesson.

What this command does Reads the local image metadata and prints only its architecture so you can confirm that the expected ENJYRA image exists.

Expected result

amd64 or arm64

If it fails

Confirm the start command completed first, then compare the image, container, port, or script name exactly with this lesson.

Reset My Lab

Reset lab

Where to runTerminalWorking directoryAny terminal folderPlatformmacOS · Linux · Windows

COMMAND

docker rm -f enjyra-v6-structured-output-05 2>/dev/null || true

docker rm -f enjyra-v6-structured-output-05 2>/dev/null || true

docker rm -f enjyra-v6-structured-output-05 2>$null

Before You Run

Your terminal is open in Any terminal folder.
Docker Desktop or Docker Engine is running.

Why You're Running This

Returns this lesson’s local runtime or resources to a known starting state without touching unrelated projects.

Command Breakdown

docker: Runs the docker tool with the shown arguments to complete this lesson step.

Expected Result

This lesson’s runtime state is reset or the named lesson container is removed.

What Changed

BeforeThe action shown by this card has not yet been completed or verified.

AfterThe command has completed and the next lesson step has the state it needs.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

Check Docker is running and make sure you are using this lesson’s exact container or launcher name.

What this command does Returns this lesson’s local runtime or resources to a known starting state without touching unrelated projects.

Expected result

This lesson’s runtime state is reset or the named lesson container is removed.

If it fails

Check Docker is running and make sure you are using this lesson’s exact container or launcher name.

Final Check

Prove What You Learned

This lesson contains 4 focused questions. You answered 2 before the build; complete the remaining 2 now.

3 / 4Interview

How should model output be treated?

Model output requires parsing, validation, and clear failures.

4 / 4Exam

The --broken run fails before field validation. Why?

Parsing must succeed before the schema can be inspected.

Skills you practised

Make AI Return Structured Data

What you'll build

Before You Build

Why use json.loads?

Why validate fields after parsing?

Mission

Understand

Build

▸Step 1Create and run the structured-data client

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

Break

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

Debug

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

Improve

Verify

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

Challenge

Ship

Next Step

Before You Start the Lab

This lesson needs

Docker setup

Before You Run

Why Do We Need Docker?

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

Lesson-specific check

Verify your setup

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

Before You Run

Why You're Running This

Command Breakdown