What Happens in the Deployment Phase of the GenAI Lifecycle

AIF-C01 By islever Editorial Last updated

Describe the ML development lifecycle.

Short answer

C During deployment the practitioner makes the trained generative AI model available in a live environment so end users can access it.

Question

What does an AI practitioner do during the deployment phase of a generative AI model lifecycle?
  1. Chooses the most useful and varied data to train the model
  2. Sets up and optimizes AWS infrastructure for the live production environment
  3. Makes the model available in a live environment for end users to access Correct Answer
  4. Continuously trains the model on new data to improve accuracy

Key Takeaways

It tests whether you can distinguish deployment itself — exposing the model to live users — from the adjacent activities of infrastructure engineering, data preparation and retraining.

This AIF-C01 question tests the deployment stage of the generative AI model lifecycle, where the practitioner makes the trained model available to end users in a live environment (option C).

Common mistake: Option B is the most tempting distractor: setting up and optimizing AWS production infrastructure supports deployment but is an MLOps platform task, not the defining deployment deliverable of model availability to end users.

Expert Analysis

The deployment boundary The ML development lifecycle in the AIF-C01 blueprint separates distinct activities: selecting and preparing training data, training, evaluation, deployment, and ongoing monitoring/retraining. The deployment phase is defined by a single act — moving the already-trained foundation model out of experimentation and into a production environment where it is reachable and consumable. Option C states exactly that: making the model available in a live environment for end users to access.

Why option B is the near-miss Option B sounds deployment-adjacent because hosting a model runs on AWS infrastructure (Amazon SageMaker endpoints, Amazon Bedrock, Lambda, API Gateway). But configuring and optimizing infrastructure is an enabling MLOps/platform task, not the defining deliverable of the deployment stage itself. When two options both touch production, pick the one describing the practitioner's outcome — model availability to end users — rather than the underlying plumbing. This is the trap the comments call out: "the primary definition of the deployment phase is the act of hosting and exposing the model for live use."

Eliminating A and D Option A, choosing the most useful and varied data, belongs to the data selection/preparation stage that precedes training. Option D, continuously training on new data to improve accuracy, describes retraining or the iterative improvement loop that follows monitoring — not deployment. A model is deployed once it is served; retraining is triggered later when monitoring reveals drift or degraded quality.

Therefore the correct choice is C.

Official Reference

https://docs.aws.amazon.com/sagemaker/latest/dg/deploy-model.html

This link opens the official vendor documentation in a new tab.

Exam Strategy

When asked to identify a lifecycle stage, anchor on the stage's defining deliverable rather than its supporting work: deployment = the model is reachable and consumable by end users. Read every option and tag it with the stage it belongs to (data prep, training, deployment, monitoring) before choosing; any option describing enabling infrastructure or later iteration is a distractor unless it directly delivers the stage's outcome.

Frequently Asked Questions

Why is setting up AWS production infrastructure not the deployment phase answer?

Infrastructure setup enables and supports deployment, but the deployment phase's defining outcome is that the trained model is exposed and accessible to end users. Infrastructure optimization is typically treated as an MLOps or platform responsibility.

Which lifecycle stage does choosing training data belong to?

Selecting and preparing the most useful, varied training data happens before training, in the data preparation or data selection stage of the ML development lifecycle, not during deployment.

When does continuous retraining on new data occur?

Continuous retraining is part of the monitoring and iterative improvement stage that follows deployment, triggered when production monitoring detects drift or a drop in model quality.

More AIF-C01 FAQ →

Community Votes

C
100%

100% of anonymous learners picked answer C. Votes are pick records left by other test-takers — they are not the verified answer.

Discussion 2 contributions

Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.

From public community threads 2

  1. NareshGondi 👍 1 Answered C
    The primary definition of the deployment phase itself is the act of hosting and exposing the model for live use.
  2. NareshGondi 👍 1 Answered C
    The primary definition of the deployment phase is the act of hosting and exposing the model for live use.

Reader comments & corrections

No reader corrections yet — spotted an error or have a note? Share it below.

Log in to comment, report an error, or add a note about this question.

How to use a custom trained model in Amazon Bedrock?

S-grade

How to Increase Image Specificity in Amazon Bedrock Stable Diffusion?

S-grade

How to Quickly Deploy and Consume Foundation Models in a VPC on AWS?

S-grade

Which AWS Service Sends Email Notifications When ISV Compliance Reports Are Available?

S-grade

Practice All AIF-C01 Questions

Access 100 questions with complete answers and detailed explanations.

View Full AIF-C01 Practice Test →

← Back to AIF-C01 Study Guide