> ## Documentation Index
> Fetch the complete documentation index at: https://docs.cloudeval.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Chat, AI Guide, or reports?

> Choose a conversation, a visual walkthrough, or a structured report for your cloud review.

Start with the result you need. These tools can support the same review, but they serve different purposes.

| Your goal                                      | Start with                         | What to do next                                                                              |
| ---------------------------------------------- | ---------------------------------- | -------------------------------------------------------------------------------------------- |
| Understand an unfamiliar project               | **AI Guide**                       | Walk through the architecture, inspect evidence, and record unanswered questions.            |
| Investigate one resource or ask a follow-up    | **AI Chat**                        | Include the selected resource or project context and ask for the evidence behind the answer. |
| Review findings, costs, or recommended actions | **Reports**                        | Check source freshness, coverage, and applicability before prioritizing work.                |
| Inspect containment or relationships yourself  | **Architecture / Dependency view** | Select resources and use the assistant when you need an explanation.                         |

## A practical review sequence

<Steps>
  <Step title="Orient with AI Guide">
    Open the project and choose **Complete architecture tour**. Identify the main resources, boundaries, and chapters with missing evidence.
  </Step>

  <Step title="Ask a specific follow-up in AI Chat">
    Select the relevant resources and ask, for example: “Which recorded relationships connect these resources, and what cannot be inferred from them?”
  </Step>

  <Step title="Check the report">
    Review the relevant cost or posture findings. Confirm that the report covers the same project source and that it is recent enough for your decision.
  </Step>

  <Step title="Verify before changing anything">
    Use the source template, cloud configuration, or operational evidence to check the recommendation. Record the next action and owner.
  </Step>
</Steps>

## What kinds of comparisons can I make?

| Comparison                                 | How to approach it                                                                                                                      |
| ------------------------------------------ | --------------------------------------------------------------------------------------------------------------------------------------- |
| Architecture versus Dependency             | Switch views of the same project: containment and overall shape versus recorded relationships.                                          |
| Two resources in one project               | Select them and ask AI Chat a specific question about their available properties or relationships. Ask it to identify missing evidence. |
| A resource versus documented guidance      | Inspect the cited guidance and verify that it applies to the configuration in your project.                                             |
| Before and after a change, or two projects | Do not assume an ordinary AI Guide is comparing snapshots. Check the currently available workflow and explicitly identify both sources. |
| Cloudeval versus another product           | This page compares Cloudeval workflows, not competitor capabilities or pricing.                                                         |

An ordinary AI Guide is a walkthrough of its loaded project context. It is not automatically a before/after review, a live traffic monitor, or proof that a recommended change has been applied. See [Interfaces and workflows](/trust/interfaces-and-workflows) for release boundaries.

## Example questions

* **AI Guide:** “Help me understand the project boundaries and major resources.”
* **AI Chat:** “What evidence supports this dependency, and which configuration should I inspect next?”
* **Reports:** “Which findings are actionable for this source, and what evidence supports each one?”

Continue with [AI Guide](/guides/results/ai-guide), [Reports](/guides/results/reports), or [AI Guide troubleshooting](/help/ai-guide).