fix issues

Author: madiepev

learn-pr/wwl-data-ai/evaluate-optimize-agents/3-git-based-experimentation-workflow.yml

@@ -1,14 +1,14 @@
 ### YamlMime:ModuleUnit
 uid: learn.wwl.evaluate-optimize-agents.git-based-experimentation-workflow
 metadata:
-  title: Apply git-based workflows to optimization experiments
-  description: Learn how to organize agent optimization experiments using git branches, systematic testing scripts, and documented evaluation results for reproducible comparisons.
+  title: Apply Git-based workflows to optimization experiments
+  description: Learn how to organize agent optimization experiments using Git branches, systematic testing scripts, and documented evaluation results for reproducible comparisons.
   ms.date: 02/17/2026
   author: madiepev
   ms.author: madiepev
   ms.topic: unit
   ai-usage: ai-generated
-title: Apply git-based workflows to optimization experiments
+title: Apply Git-based workflows to optimization experiments
 durationInMinutes: 8
 content: |
   [!include[](includes/3-git-based-experimentation-workflow.md)]

learn-pr/wwl-data-ai/evaluate-optimize-agents/6-knowledge-check.yml

@@ -3,7 +3,7 @@ uid: learn.wwl.evaluate-optimize-agents.knowledge-check
 title: Knowledge check
 metadata:
   title: Knowledge check
-  description: "Test your understanding of agent evaluation experiments, git-based workflows, and evaluation rubrics."
+  description: "Test your understanding of agent evaluation experiments, Git-based workflows, and evaluation rubrics."
   ms.date: 02/17/2026
   author: madiepev
   ms.author: madiepev
@@ -17,11 +17,11 @@ content: |
 quiz:
   title: "Check your knowledge"
   questions:
-  - content: "What is the primary reason for organizing agent optimization experiments into separate git branches?"
+  - content: "What is the primary reason for organizing agent optimization experiments into separate Git branches?"
     choices:
     - content: "To enable parallel development by multiple team members simultaneously"
       isCorrect: false
-      explanation: "Incorrect: While git supports parallel development, the primary reason for separate experiment branches is controlled comparison, not collaboration."
+      explanation: "Incorrect: While Git supports parallel development, the primary reason for separate experiment branches is controlled comparison, not collaboration."
     - content: "To isolate specific changes and attribute performance differences to individual modifications"
       isCorrect: true
       explanation: "Correct: Separate branches enable controlled comparison by testing one change at a time, making it clear which modification caused observed performance differences."

learn-pr/wwl-data-ai/evaluate-optimize-agents/7-summary.yml

@@ -2,7 +2,7 @@
 uid: learn.wwl.evaluate-optimize-agents.summary
 metadata:
   title: Summary
-  description: Summary of key takeaways for structured agent optimization including evaluation metrics, git-based workflows, consistent scoring rubrics, and evidence-based decision making.
+  description: Summary of key takeaways for structured agent optimization including evaluation metrics, Git-based workflows, consistent scoring rubrics, and evidence-based decision making.
   ms.date: 02/17/2026
   author: madiepev
   ms.author: madiepev

learn-pr/wwl-data-ai/evaluate-optimize-agents/includes/2-design-evaluation-experiments.md

@@ -65,21 +65,21 @@ Representative test prompts cover the spectrum of real-world usage. For the Adve
 
 - **Digital nomads planning weekend hikes**: "I'm hiking in the Scottish Highlands in March, what waterproof gear do I need from Adventure Works?"
 - **Families preparing for their first outdoor adventure**: "We're taking our teenagers on easy trails near London next month, what basic equipment should we buy or rent?"
-- **Experienced hikers planning extended trips**: "I need a complete gear list for five-day backpacking trip in moderate terrain with variable weather"
+- **Experienced hikers planning extended trips**: "I need a complete gear list for five-day backpacking trip in moderate terrain with variable weather."
 
 Edge cases test how the agent handles challenging situations:
 
 - **Ambiguous requests**: "What should I pack for hiking?"
-- **Incomplete trip details**: "I need gear for Scotland"
+- **Incomplete trip details**: "I need gear for Scotland."
 - **Last-minute gear changes**: "Can I swap my camping equipment rental for different sizes?"
 
 Including five to 10 diverse test prompts provides sufficient coverage for manual testing and smoke tests while remaining practical for human evaluation. Each test prompt captures the user query, expected information needs, and ideal response characteristics.
 
 Success criteria establish what constitutes acceptable performance before you run experiments. Setting thresholds in advance prevents rationalizing disappointing results. Adventure Works defines success thresholds across all three optimization dimensions:
 
-- **Quality**: Average 4.2+ (five-point scale), minimum 3.5 per response to align with customer satisfaction targets and prevent trust erosion
-- **Cost**: 60% expense reduction to achieve operational efficiency goals while maintaining 85% resolution rate
-- **Performance**: Average response time <30 seconds, time-to-first-token <2 seconds (streaming) to ensure acceptable user experience for real-time interactions
+- **Quality**: Average 4.2+ (five-point scale), minimum 3.5 per response to align with customer satisfaction targets and prevent trust erosion.
+- **Cost**: 60% expense reduction to achieve operational efficiency goals while maintaining 85% resolution rate.
+- **Performance**: Average response time <30 seconds, time-to-first-token <2 seconds (streaming) to ensure acceptable user experience for real-time interactions.
 
 Business requirements influence these thresholds: customer-facing agents handling trip planning need higher quality standards and faster response times than internal tools.
 

learn-pr/wwl-data-ai/evaluate-optimize-agents/includes/3-git-based-experimentation-workflow.md

@@ -91,7 +91,7 @@ For the Adventure Works experiments, you might document your comparison:
 | prompt-v2-concise | Maintains quality, more focused | Yes (4.4 avg) |
 | gpt4o-mini-model | Lower quality on complex prompts | No (4.1 avg, below 4.2 threshold) |
 
-If `prompt-v2-concise` meets your quality threshold and improves conciseness, use git to merge the winning experiment:
+If `prompt-v2-concise` meets your quality threshold and improves conciseness, use Git to merge the winning experiment:
 
 ```bash
 git checkout main
@@ -102,4 +102,4 @@ git push origin main --tags
 
 For experiments that don't meet criteria, document why before deciding whether to keep or delete the branch: "gpt4o-mini-model: Quality dropped below 4.2 threshold on complex trip planning prompts. Not recommended for production."
 
-With git workflows established for organizing experiments, you're ready to execute the actual evaluations by running agents against test prompts and systematically scoring the results.
+With Git workflows established for organizing experiments, you're ready to execute the actual evaluations by running agents against test prompts and systematically scoring the results.

learn-pr/wwl-data-ai/evaluate-optimize-agents/includes/5-exercise.md

@@ -3,7 +3,7 @@ In this exercise, you evaluate two prompt versions of the Trail Guide Agent and
 
 Throughout this exercise, you:
 
-- Design evaluation experiments using git branches to manage and compare prompt versions
+- Design evaluation experiments using Git branches to manage and compare prompt versions
 - Manually evaluate AI agent responses against structured quality criteria (Intent Resolution, Relevance, Groundedness)
 - Compare agent versions across test scenarios to identify quality differences and cost trade-offs
 

learn-pr/wwl-data-ai/evaluate-optimize-agents/includes/7-summary.md

@@ -5,7 +5,7 @@ You've learned how to optimize AI agents through structured evaluation that tran
 
 Effective optimization depends on clear metrics that measure quality, cost, and performance. Quality metrics like Intent Resolution, Relevance, and Groundedness reveal whether agents serve user needs effectively. Cost metrics quantify token usage and operational expenses, enabling you to calculate the financial impact of model changes. Performance metrics measure response times that directly affect user experience. Together, these metrics provide objective criteria for comparing agent variants.
 
-## Organize experiments with git-based workflows
+## Organize experiments with Git-based workflows
 
 Git-based workflows bring engineering discipline to agent optimization. You create one branch per experiment variant, isolating specific changes like prompt modifications or model switches. Each branch maintains test prompts, evaluation scripts, and documented results. This structured approach lets you test changes safely, compare experiments systematically, and merge successful optimizations to production with confidence.
 

learn-pr/wwl-data-ai/evaluate-optimize-agents/index.yml

@@ -2,7 +2,7 @@
 uid: learn.wwl.evaluate-optimize-agents
 metadata:
   title: Evaluate and optimize AI agents through structured experiments
-  description: Learn how to evaluate and optimize AI agents systematically through structured experiments that measure quality, cost, and performance. Design evaluation metrics, apply git-based workflows, create consistent scoring rubrics, and make evidence-based optimization decisions.
+  description: Learn how to evaluate and optimize AI agents systematically through structured experiments that measure quality, cost, and performance. Design evaluation metrics, apply Git-based workflows, create consistent scoring rubrics, and make evidence-based optimization decisions.
   ms.date: 02/17/2026
   author: madiepev
   ms.author: madiepev
@@ -11,17 +11,17 @@ metadata:
   ms.service: azure-ai-foundry
 title: Evaluate and optimize AI agents through structured experiments
 summary: |
-  Learn how to optimize AI agents through structured evaluation that transforms guesswork into evidence-based engineering decisions. You'll explore how to design evaluation experiments with clear metrics for quality, cost, and performance; organize experiments using git-based workflows; create evaluation rubrics for consistent scoring; and compare results to make informed optimization decisions.
+  Learn how to optimize AI agents through structured evaluation that transforms guesswork into evidence-based engineering decisions. You'll explore how to design evaluation experiments with clear metrics for quality, cost, and performance; organize experiments using Git-based workflows; create evaluation rubrics for consistent scoring; and compare results to make informed optimization decisions.
 abstract: |
   In this module, you:
   - Design evaluation experiments with clear metrics for quality, cost, and performance
-  - Apply git-based workflows to organize and compare agent variants systematically
+  - Apply Git-based workflows to organize and compare agent variants systematically
   - Create evaluation rubrics that ensure consistent scoring across human evaluators
   - Compare experiment results to make evidence-based optimization decisions
 prerequisites: |
   Before starting this module, you should have:
   - Basic understanding of AI agents and large language models
-  - Familiarity with git version control workflows
+  - Familiarity with Git version control workflows
   - Experience with Microsoft Azure AI Foundry or similar AI development platforms
 iconUrl: /training/achievements/generic-badge.svg
 levels: