> For the complete documentation index, see [llms.txt](https://docs.duplocloud.com/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.duplocloud.com/docs/advanced-topics/analytics.md).

# Analytics

The **Analytics** section covers two admin-only screens: the **Platform Analytics** dashboard (usage/cost/ticket metrics across the whole platform) and **Observability** (LangFuse agent traces, embedded in the portal and searchable by session).

{% hint style="info" %}
This is distinct from [AI Dashboards](https://docs.duplocloud.com/docs/advanced-topics/ai-dashboards.md), which are agent-built dashboards over your connected cloud/infra providers. Platform Analytics is a fixed, built-in view over DuploCloud's own usage data — it isn't agent-generated and isn't scoped to a provider.
{% endhint %}

## Platform Analytics Dashboard

Go to **AI Admin → Analytics → Platform Analytics**. The **Overview** tab shows platform-wide totals — users, active users in the last 30 days, groups, tickets, projects, workspaces, providers, skills, and personas.

![Platform Analytics overview](https://2471407984-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F68cb0s9ce5UIUKWPuYs8%2Fuploads%2Fgit-blob-8f83d63c3cab5b8c283ab81d2209ee6fd7af45aa%2Fanalytics-step-01-overview.png?alt=media)

The remaining tabs break these totals down further:

* **Tickets** — status distribution (Open/In Progress/Closed), top workspaces by ticket volume, tickets by user, and tickets by scope

![Tickets tab](https://2471407984-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F68cb0s9ce5UIUKWPuYs8%2Fuploads%2Fgit-blob-c4fbae6d21aaa67b48302eea91897abf5b3e08be%2Fanalytics-step-02-tickets-tab.png?alt=media)

* **Cost** — total and average cost, the top workspace by spend, and cost broken down by workspace, by user, and by scope

![Cost tab](https://2471407984-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F68cb0s9ce5UIUKWPuYs8%2Fuploads%2Fgit-blob-9116cc98280779788f01ee3243442aa367fd026a%2Fanalytics-step-03-cost-tab.png?alt=media)

* **Quotas** — how many quota definitions and mappings exist, then a card for each [quota mapping](https://docs.duplocloud.com/docs/advanced-topics/access-control/quotas.md#quota-mappings) showing its limit and period, what it applies to, and current spend (or average spend per user, ticket, or workspace) against the limit, with a badge showing whether it's within the limit. The cards label a mapping's Context as **Scope** (System appears as **Platform**) and its Apply To as **Type**.

![Quotas tab](https://2471407984-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F68cb0s9ce5UIUKWPuYs8%2Fuploads%2Fgit-blob-2d63add842a333bcecabe29ae2cc8bb271abb6ef%2Fanalytics-step-04-quotas-tab.png?alt=media)

* **Projects** — project counts by workspace, by scope, and by user

![Projects tab](https://2471407984-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F68cb0s9ce5UIUKWPuYs8%2Fuploads%2Fgit-blob-932a69077db38d7c25c41f38dfcb1eae9cdc46bc%2Fanalytics-step-05-projects-tab.png?alt=media)

* **Users** — active vs. inactive users (based on login in the last 30 days), plus breakdowns by group and by permission set

![Users tab](https://2471407984-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F68cb0s9ce5UIUKWPuYs8%2Fuploads%2Fgit-blob-034ea0c7e2bc14b54eca0753058351d98ffdb323%2Fanalytics-step-06-users-tab.png?alt=media)

{% hint style="info" %}
Nothing needs to be configured for this dashboard — it's computed from data that already exists (workspaces, tickets, users, quotas, and so on), so it populates automatically as your platform is used.
{% endhint %}

## Observability

The **Observability** pages show agent traces recorded in [LangFuse](https://langfuse.com), an open-source LLM observability tool. Each trace is tagged with the originating ticket, workspace, user, and model, so a trace can be followed back to the exact ticket that produced it.

### Set up LangFuse

DuploCloud can deploy LangFuse for you as part of the AI HelpDesk Helm chart. It's off by default; enabling it (`langfuse-stack.enabled`) deploys:

* LangFuse itself, on its own hostname, storing traces in the same ClickHouse database as [Audit Trail](https://docs.duplocloud.com/docs/concepts/helpdesk/audit-trail.md)
* A DuploCloud sign-in proxy in front of it, so admins open LangFuse with their portal session instead of a separate LangFuse login. By default only the **Administrator** role is let through, and everyone signs in to LangFuse as one shared account (**Duplo SSO**)
* The agent runtime's LangFuse credentials, so traces start flowing without further setup

This is a deployment-level step, done in the Helm values rather than from the portal.

You can also use a LangFuse instance you run yourself. It must allow the portal to embed it in an iframe (LangFuse Cloud does not), and the agent runtime needs that instance's API keys for traces to reach it.

### LangFuse

Go to **AI Admin → Analytics → Observability → LangFuse**. The page embeds the LangFuse UI, so you can browse projects, traces, and sessions without leaving the portal.

![LangFuse embedded in the portal](https://2471407984-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F68cb0s9ce5UIUKWPuYs8%2Fuploads%2Fgit-blob-371c7313f1e9d28a0d513c75352d516b05c33832%2Fanalytics-step-07-langfuse.png?alt=media)

The page shows the URL saved in the `langfuse-url` [System Setting](https://docs.duplocloud.com/docs/advanced-topics/settings.md#system-settings). Set it to your LangFuse hostname — enabling LangFuse in the Helm chart doesn't fill it in. Until it's set, the page shows an empty state with a **Configure LangFuse URL** button that opens the Add System Setting form with `langfuse-url` already selected.

### LangFuse Sessions

Go to **AI Admin → Analytics → Observability → LangFuse Sessions** to search agent sessions without opening LangFuse.

![LangFuse Sessions](https://2471407984-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F68cb0s9ce5UIUKWPuYs8%2Fuploads%2Fgit-blob-40cd6f5ce19d52a8dc08f96431556e6ec3a81508%2Fanalytics-step-08-langfuse-sessions.png?alt=media)

Filter by **Workspace**, **User**, and **Search text** (a session or ticket ID), choose a **Time Range** (**Last 24 hours**, **Last 7 days**, **Last 30 days**, or **Custom**), and click **Search**. Each row shows the session's workspace, user, start and last-activity times, number of traces, total cost, and total latency. The link icon at the end of a row opens that session in LangFuse; it appears only when `langfuse-url` is set.

This page reads session data straight from the ClickHouse database LangFuse writes to. If the backend has no connection to that database, the page says session search isn't enabled for the environment — that is not the same as no matching sessions.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.duplocloud.com/docs/advanced-topics/analytics.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
