Managed online endpoints with traffic splitting are the best fit because Azure Machine Learning supports multiple deployments behind a single online endpoint and allows production traffic to be distributed between them. This enables a blue-green or progressive rollout strategy : deploy the new model version alongside the current version, validate it independently, send a small percentage of production traffic to it, monitor behavior, and then gradually increase traffic. If problems occur, traffic can be redirected immediately to the previous deployment, providing a practical rollback mechanism.

This approach also aligns with Fabrikam’s requirement to favor managed services and automation because the team has limited Azure DevOps experience. Managed online endpoints handle much of the serving, scaling, security, and monitoring infrastructure automatically and support both traffic splitting and mirrored traffic for safer validation.

Option B can also implement blue-green deployment, but AKS introduces substantially more infrastructure and operational management than Fabrikam requires. Option A would require custom deployment, routing, scaling, and rollback logic. Option D is intended for asynchronous batch inference rather than staged real-time production serving.

Study Guide Reference: Implement machine learning model lifecycle and operations — managed online endpoints, multiple deployments, traffic splitting, blue-green deployment, staged rollout, monitoring, and rollback.

In Microsoft Foundry ' s built-in evaluation framework, each evaluator measures a specific dimension of language quality. Fluency evaluates the grammatical correctness, sentence structure, and overall linguistic smoothness of generated text. A high fluency score means the response reads naturally, with proper grammar, punctuation, and sentence construction — regardless of factual accuracy or relevance to the topic. Coherence (option A) measures whether ideas flow logically from one sentence to the next — it is about logical structure, not grammar. Textual Similarity (option B) measures how closely the generated text matches a reference text using metrics like ROUGE or BLEU — it is a comparison metric, not a standalone quality dimension. Groundedness (option C) measures whether claims are supported by the provided context — it is a factual fidelity metric. For grammatical correctness specifically, Fluency is the correct evaluator.

Microsoft Learn Reference Topic: Built-in AI evaluators in Microsoft Foundry – Fluency, Coherence, Groundedness, and Similarity

Different query types require fundamentally different retrieval algorithms. For queries that include exact policy identifiers, keyword or BM25 search is the correct choice because BM25 scores documents based on term frequency and inverse document frequency — an exact match on a specific policy identifier such as POL-2024-HR-042 scores very highly regardless of semantic context. This is the right approach when semantic similarity is low but exact term matching is critical. For natural-language questions where keywords may not be an exact match, semantic or vector search is the correct choice because vector embeddings capture meaning rather than exact tokens, finding relevant documents even when the user ' s vocabulary differs from the document ' s terminology. Azure AI Search supports both modes through its hybrid search capability, and the correct configuration maps each query type to its optimal retrieval algorithm.

Microsoft Learn Reference Topic: Configure hybrid search in Azure AI Search – BM25 keyword search vs. semantic vector search

Chunk size and overlap directly affect the precision and completeness of information retrieved in a RAG architecture. Microsoft identifies effective chunking as a critical part of successful RAG implementation. If chunks are too large, they can contain excessive unrelated information, reducing retrieval precision because the vector representation covers multiple concepts. Conversely, chunks that are too small can omit necessary surrounding context or separate related information across chunk boundaries, resulting in incomplete retrieval.

Overlap helps preserve semantic continuity between adjacent chunks. Microsoft recommends using overlap particularly with fixed-size chunking because relevant facts or concepts can span chunk boundaries. The optimal chunk size and overlap should therefore be evaluated against representative documents and queries rather than treated as fixed constants. Microsoft specifically notes that larger chunks can preserve contextual richness, while appropriately configured overlap prevents contextual information from being lost at segmentation boundaries.

Temperature controls generation variability , not document retrieval. Token limits constrain how much content a model processes but do not inherently make retrieval more relevant. Embedding strategy can also influence retrieval quality, but given the stated symptom of irrelevant or incomplete retrieved passages , tuning chunk size and overlap is the directly targeted remediation.

Study Guide Reference: Optimize generative AI systems and model performance — RAG optimization, document chunking, retrieval precision, contextual completeness, and grounding.

===============

All three requirements point to Git as the underlying mechanism, but each uses a different Git capability. For reviewing changes before production: a branch-based workflow where prompt changes are made on feature branches and merged to the main branch only after pull request approval enforces the review gate. For restoring previous versions: Git ' s commit history and tag system provide a precise, immutable record of every prompt state, and a git revert or checkout to a specific commit SHA restores any prior version instantly. For governance parity with application code: by storing prompts in the same Git repository as application code, all the same CI/CD, branch protection, code review, and audit trail policies apply automatically. The alternatives such as Blob Storage or embedded configuration files lack native review workflows, branch protection, and full audit history.

Microsoft Learn Reference Topic: Prompt management and versioning with Git integration in Microsoft Foundry

The key concept here is low-priority (spot) VMs, which are available at a significant discount because Azure can reclaim them at any time. Azure Machine Learning compute clusters are the only target in the list that directly supports low-priority VM nodes as a cost-saving configuration. You set the minimum and maximum node counts and specify that new nodes should be provisioned as low-priority. Azure Container Instances (ACI) does not support low-priority pricing. Local deployment runs on the developer ' s machine with no Azure billing model. Azure Kubernetes Service (AKS) does support spot node pools but requires significantly more infrastructure management and is not the primary mechanism for low-priority compute in Azure Machine Learning. The exam tests whether you know that AML compute clusters are the managed way to leverage low-priority discounts inside Azure Machine Learning.

Microsoft Learn Reference Topic: Create and manage Azure Machine Learning compute clusters – Low-priority VMs

For tracking changes across contributors, Git integration is the answer: by connecting the Microsoft Foundry project to a Git repository, every prompt file change is tracked as a commit with author attribution, timestamp, and diff view, and pull requests enforce review before changes reach production. The Git history provides the complete audit trail and rollback capability needed for traceability. For allowing applications to consume updated prompts without requiring redeployment, Microsoft Foundry ' s prompt management feature allows prompts to be stored and versioned as named artifacts in the project. Applications reference prompts by name and load the latest approved version at inference time, rather than having prompt text hard-coded in the application deployment artifact. This decoupling means updating a prompt is a content operation — not a code deployment — so applications automatically pick up the new prompt without any redeployment.

Microsoft Learn Reference Topic: Prompt management in Microsoft Azure AI Foundry – Git integration and dynamic prompt versioning