Monitoring AI Usage and Request History

Opening the AI usage dashboard and understanding what it measures

In Atloria, start from the authenticated workspace and open the Admin area. On the main Admin screen, look for the Analytics card and open it to reach Analytics & Insights. This screen is presented as the place for Usage statistics and performance metrics, so it is the natural starting point when you want a high-level view of AI activity across your organization. [SCREENSHOT: Admin screen showing the Analytics card]

1. Open Admin from the main workspace navigation.
2. Select Analytics.
3. Review the Analytics & Insights header to confirm you are in the reporting area.
4. Look for summary sections that show overall activity for the selected period, such as total requests, usage totals, and recent activity when available.

When you work with AI usage reporting, it helps to separate two kinds of information:

• Dashboard totals and charts show the big picture for a selected time period.
• Request history shows one AI interaction at a time.

Use the dashboard when you want to answer questions like:

• Which team is using Atloria AI features most?
• Did usage increase this week?
• Which workspace or feature is driving the largest share of activity?

Open request history when you need to inspect a specific interaction, such as a failed request, an unusually expensive request, or a response that looked incorrect.

If a date range picker is available on your AI usage screen, changing it updates the totals and charts together. Common presets usually include:

• Today
• Last 7 days
• Current month

These presets are useful for different review habits. Today helps with live monitoring, Last 7 days helps spot short-term changes, and Current month helps with budget and team reporting. Stay on the dashboard when you need trends and totals. Move into request history when you need evidence behind those numbers.

Filtering usage data to review teams, users, and time periods

Filtering is what turns a broad AI usage report into something you can act on. In Atloria, once you are in the reporting area, use the filter controls shown on the page to narrow the results to the people, projects, or time periods you want to review. Depending on your access level, you may be able to filter by team, project, user, model, or request status.

1. Open the AI usage reporting area from Admin.
2. Set the Date range first so every total and chart reflects the same period.
3. Apply a team, project, or user filter to reduce the results to the group you want to review.
4. If available, add a model or status filter to focus on a specific type of AI request.
5. Review the updated totals, charts, and lists after each filter change.

A good habit is to change one filter at a time. That makes it easier to understand why the numbers changed. For example, if overall usage looks high, switch from Current month to Last 7 days and then narrow the view to one team. This helps you tell the difference between a normal monthly total and a short-term spike.

Search fields are especially useful when a support lead needs to isolate one person’s activity. Enter the user’s name or email in the search box, then combine that with the date range to find the exact period in question. If Atloria offers saved filter combinations in your workspace, use them for recurring reviews such as:

• Weekly team usage review
• Monthly cost review
• Recent failed requests
• Activity for one project workspace

Sorting helps when the list is long. Sort by:

• Highest usage to find cost drivers
• Most requests to find heavy activity
• Latest activity to investigate a recent issue

Use filters for comparison as well. Review one time window, note the totals, then switch to another window and compare the trend chart or grouped totals. This is often the fastest way to spot unusual growth or a sudden drop in request volume.

Inspecting request history for prompts, responses, and outcomes

When the dashboard tells you something changed, request history helps you find out why. In Atloria, open the request history list from the AI usage area and review the records line by line. This view is meant for auditing and troubleshooting, so it focuses on individual interactions instead of overall totals. [SCREENSHOT: Request history list with columns for time, user, status, and usage]

1. Open the AI usage area from Admin.
2. Switch from the summary view to the Request history list.
3. Use the search box or filters to narrow the list before opening individual records.
4. Click a request row to open its detail panel or detail page.
5. Review the prompt, response, status, timing, and usage values for that request.

The request list commonly includes fields like these:

Once you open a request, focus on three parts of the record:

• The submitted prompt
• The generated response
• The request details around it, such as status, time, and feature source

Status values matter during investigations:

• Success means the request completed normally.
• Error means it failed and needs review.
• Timeout means it took too long and did not finish in time.
• Blocked means it was stopped by a rule or restriction.
• Retried means Atloria attempted the request again after a problem.

Use pagination when the list spans many pages, especially for busy teams. If you cannot find a request quickly, combine search, date range, and status filters before moving through pages manually. That approach is much faster than scanning the full history.

Using history records to manage cost, quality, and oversight

AI usage records are most valuable when you use them to make decisions. In Atloria, request history is not just a log of past activity. It helps administrators and team leads control spending, improve output quality, and maintain accountability across teams and workspaces.

1. Start with the dashboard to identify a spike, trend, or unusually active group.
2. Open request history for the same date range and filters.
3. Sort or filter by the highest cost, highest usage, or repeated failures.
4. Open individual records that stand out.
5. Use those records to decide whether the issue is cost-related, quality-related, or operational.

For cost review, look at per-request usage values such as Cost or Token usage, along with the Model and Feature columns. Expensive patterns often appear as repeated high-usage requests, a sudden increase from one team, or a workflow that uses a more costly model more often than expected. That is the point where an administrator may decide to tighten usage rules or review AI settings with the team responsible.

For quality review, open the request details and compare the prompt with the response. Team leads often look for:

• Repeated retries for the same task
• Low-quality or off-topic responses
• Prompts that are too vague to produce reliable results
• Misuse patterns, such as unnecessary repeated generation

For oversight, the most important fields are Timestamp, User, Team, Feature, and Status. Together, these create an activity trail that shows who did what and when. That makes request history useful during internal reviews, support follow-up, and audit preparation.

Use aggregate dashboard trends for planning questions, such as whether usage is growing month over month. Use individual request records when you need to understand a specific incident, review one user’s activity, or confirm what happened during a support case.

Exporting and sharing usage records for reporting and audits

When you need to report on AI activity outside the screen, use the export option from the usage dashboard or request history view. In Atloria, exports are most useful after you have already narrowed the data to the exact scope you want. That way, the file matches what you are reviewing on screen and does not include unrelated records.

1. Open the AI usage dashboard or Request history.
2. Apply the exact Date range, team, project, user, model, and status filters you need.
3. Confirm the visible totals or list count before exporting.
4. Select the Export action.
5. Download the file and review the first few rows to confirm it matches the filtered view.

Before exporting, preserve the current filter state as carefully as possible. If Atloria provides visible filter chips, selected dropdown values, or a saved view, verify those settings before you click Export. This is especially important for finance reviews and audit requests, where a small filter difference can change the totals significantly.

For most reporting needs, include fields like these in the export:

Be careful when sharing exports that include prompt and response content. Those records may contain sensitive operational details or customer-related text. Share only the fields needed for the review. For example:

• Finance usually needs usage totals, timestamps, teams, and cost-related values.
• Support teams often need status, duration, feature, and selected prompt or response examples.
• Compliance reviews may need user, timestamp, status, and full activity scope.

If you need broader administrative reporting, see Monitoring Administrative Analytics and Activity.

Resolving common issues when usage totals or request logs do not look right

If AI usage numbers or request logs seem wrong in Atloria, the cause is usually a filter mismatch, a scope issue, or a visibility setting. Start by checking what is selected on screen before assuming the underlying activity is missing.

1. Recheck the Date range and make sure it matches the period you intend to review.
2. Confirm the team, project, workspace, or user filters currently applied.
3. Verify whether the view includes all models and all statuses, or only a subset.
4. Compare the dashboard view and the request history view using the same filters.
5. If needed, repeat the search with broader filters to rule out an overly narrow view.

If totals seem too high or too low, the first thing to inspect is the date filter. A Current month view can look much larger than Last 7 days, even when activity is normal. Next, check whether you are looking at all teams or just one workspace. A model filter can also change the numbers sharply if only one AI option is included.

If a request appears to be missing from history, work through these checks:

• Confirm you are searching under the correct user or team scope
• Remove extra search terms and try a broader search
• Check additional pages in the list
• Review whether the request may have failed before a full record was saved

If prompt or response details are not visible, the issue may be related to access level or content visibility rules. In that case, check whether your role allows you to view full request content, or whether those details are limited in your organization’s AI review process.

If an export does not match the dashboard, compare these items before exporting again:

• Date range
• Team or user filters
• Model and status filters
• Time zone used for the report
• Current sort or grouping choices

For related administrative review areas, see Reviewing Security and Audit Controls.

Overview

This document focuses on how to monitor AI activity in Atloria from a reporting and oversight perspective. The main tasks covered here are opening the Admin reporting area, reviewing high-level usage information, narrowing results with filters, inspecting individual request records, and exporting the results for reporting or audit work.

The most important idea is the difference between summary reporting and request-level inspection. Use the dashboard view when you need totals, trends, or comparisons across time periods, teams, or workspaces. Use Request history when you need to understand one interaction in detail, such as a failed request, an expensive request, or a response that needs quality review.

This guide is especially useful for people who oversee AI usage rather than create AI content directly, including:

• Administrators reviewing organization-wide activity
• Team leads checking usage patterns for their teams
• Support leads investigating failed or unusual requests
• Operations reviewers preparing usage reports or audit evidence

In Atloria, the reporting workflow usually follows a simple pattern:

1. Open the reporting area from Admin.
2. Set the date range.
3. Apply filters for the team, project, user, or status you want to review.
4. Read the totals and trends.
5. Open request history for the records behind those numbers.
6. Export the filtered results if you need to share them.

Because the Analytics & Insights area is currently positioned as the place for usage statistics and performance metrics, you should treat it as the starting point for AI usage monitoring. If you need broader admin navigation help before working in this area, read Using the Admin Workspace.

The next document in this section is Managing AI Usage and Request History, which continues from monitoring into day-to-day control and follow-up actions.

Prerequisites

Before you monitor AI usage and request history in Atloria, make sure you have the right access and enough context to interpret what you are seeing.

• You must be able to sign in to Atloria and reach the authenticated workspace.
• You need access to the Admin area or another management view that exposes AI usage reporting.
• You should know which team, project, or workspace you are responsible for reviewing.
• You should have a clear reporting period in mind, such as Today, Last 7 days, or Current month.
• You should understand whether you are looking for a trend, a specific incident, or a user-level activity trail.

It also helps to know what kind of review you are performing:

• Cost review focuses on usage totals, model choice, and per-request cost or token values.
• Quality review focuses on prompts, responses, retries, and failed outcomes.
• Operational oversight focuses on timestamps, users, teams, and request status.

If you are new to Atloria account access or need help getting into the workspace first, use these related guides:

• Accessing and Registering Your Atloria Account
• Signing In to Atloria and Solving Access Problems
• Understanding Account Entry Points and Session Navigation

If you expect to review administrative reporting regularly, you may also want to be familiar with:

• Using the Admin Workspace
• Managing User Access and Administrative Permissions
• Configuring AI Settings for Your Organization

When these prerequisites are in place, you can move through AI usage monitoring quickly and interpret the results with much more confidence.