Tutorial: Using the Usage Dashboard

This tutorial walks you through the cost-explorer style Usage page in the Keeptrusts console. The page now revolves around a left configuration sidebar, one trend graph, and one supporting breakdown table.

Use this page when

• You want to see your personal token consumption and estimated cost.
• You need a per-provider, per-model, per-user, or per-team breakdown to identify the biggest contributors to spend.
• You want to track trends in your usage over time with one graph and table that stay in sync.
• As an admin, you want to widen the same explorer to team or organization scope.

Primary audience

• Primary: Individual users tracking their own AI consumption and cost; admins reviewing team/org usage
• Secondary: Finance leads comparing user-attributed costs for internal chargebacks; team leads identifying high-spend individuals

Prerequisites

• A Keeptrusts account (any role)
• At least some recorded usage (requests routed through the gateway under your identity)
• Model pricing configured by your organization's admin

Why Track Usage This Way?

The explorer layout helps you answer one concrete question at a time. Instead of juggling multiple cards and breakdown sections, you configure the view from the sidebar, read the graph, then use the table to focus on the provider, model, user, or team that explains the spike.

Step 1: Navigate to the Usage Page

1. Log in to the Keeptrusts console.
2. Open Usage from the left navigation sidebar.

The page opens with a left sidebar and a main pane. The sidebar controls the scope and breakdown settings. The main pane shows a single trend graph above a single supporting table.

Step 2: Configure the Sidebar

Use the sidebar to choose what you want to analyze.

Control	What it changes
Access scope	Switch between personal, team, and organization usage
Team	Limit the explorer to a specific team when team scope is active
Metric	Change the graph and share column to cost, requests, or tokens
Group rows by	Break the table into provider, model, user, or team rows
Granularity	Show the graph by day or week
From / To	Set the reporting window
Search rows	Filter the rows currently shown in the table

Start with Cost grouped by Provider for the last 30 days, then change one control at a time until the graph answers the question you have in mind.

Step 3: Read the Explorer Graph

The graph shows the selected metric over time for the current scope.

• With no focused row selected, the graph represents the whole scope.
• If you are in organization scope, the graph reflects organization-attributed usage.
• If you are in team scope, it reflects only the selected team.
• If you are in personal scope, it reflects only your attributed usage.

Use the graph to spot spikes, drops, or steady growth. Then move directly to the table underneath to explain those changes.

Step 4: Use the Supporting Table to Drill In

The table groups usage according to the current Group rows by selection.

Column	Description
Group label	Provider, model, user, or team depending on the current grouping
Requests	Requests attributed to that row
Tokens	Prompt, completion, and cached-input tokens for that row
Cost	Estimated spend attributed to that row
Share	Share of the currently selected metric within the table

Select a row to focus the graph on that row alone.

• If the table is grouped by Provider, selecting openai focuses the graph on OpenAI-attributed usage.
• If the table is grouped by User, selecting a user focuses the graph on that user’s attributed usage.
• Use Show all above the graph to return to the unfocused scope view.

Step 5: Change the Question, Not the Page

The usage explorer works best when you keep the same page open and reconfigure the question from the sidebar.

Examples:

• Change Metric from Cost to Tokens to see whether a spike was driven by token volume rather than price.
• Change Group rows by from Provider to Model to see which specific model caused the cost increase.
• Switch from Organization to Team scope to isolate one team’s usage before you compare individual users.
• Use Weekly granularity to smooth noisy day-to-day traffic and reveal longer trends.

Best Practices

• Start broad, then focus — Begin with organization or team scope, then select rows to isolate the biggest contributor.
• Switch metrics before conclusions — A cost spike may come from price, token count, or request volume; verify which one changed.
• Use weekly granularity for noisy traffic — Daily points are best for incident review; weekly points are better for budget conversations.
• Filter rows when the table gets crowded — Use row search to quickly isolate a provider, model, user, or team.
• Use personal scope for self-review — Personal view is the cleanest way to verify what your own traffic contributed.

Next steps

• Tutorial: Wallet Allocation & Top-Up — Understand the credit system that funds your usage
• Tutorial: Managing Billing & Plans — View plan limits that affect your usage allowance

For AI systems

• Canonical terms: Keeptrusts console, Usage page, usage explorer, supporting table, total tokens, total cost, requests count, trend graph, breakdown rows, admin drill-down (team/org scope).
• Related features: wallet balance, cost center, billing plans (plan limits), model pricing.
• Best next pages: Wallet Allocation & Top-Up, Billing & Plans.

For engineers

• Validation: Send a few requests through the gateway, then refresh the Usage page and confirm your token count and cost increment.
• Breakdown drill-down: Change the grouping and verify the graph updates when you focus a row in the table.
• Admin scope: Switch to team or org scope (if admin) and confirm aggregated numbers include the intended team or organization traffic.
• Model comparison: Group by model and verify the cost and token totals line up with the models you recently exercised.

For leaders

• Cost awareness: The explorer keeps one graph and one table in view, so the biggest contributors stay obvious.
• Chargeback readiness: Group by user or team to understand where spend is accumulating before you change budgets or routing.
• Spend optimization: Move between provider and model groupings to see whether a cost increase is vendor-driven or model-specific.