Getting started

Dashboard

The dashboard at /dashboard is the screen you land on after signing in. It answers two questions: what is the fastest next thing to do, and what has been happening in your workspace.

Quick actions

Four tiles at the top cover the main entry points into the platform:

Open the Playground
Opens the Chat Playground to chat with any model and prototype an agent instantly.
Build a Standalone Agent
Opens the Agent Builder with the new-agent form ready.
Design a Swarm
Opens a blank Swarm Canvas to wire agents together into a multi-agent graph.
Open BI Workspace
Opens the BI Workspace, where AI builds dashboards over your connected data.

A curated row of four multi-agent templates that open directly on the canvas. They are chosen for being graphs worth reading — routing, parallel fan-out, retrieval and approval gates — and they need no setup, because their retrieval nodes point at the sample knowledge base bundled with the platform.

TemplateWhat it demonstrates
Support CopilotRouter → KB retrieval with reranking → grounded answer → LLM judge → human approval. The most complete of the four.
Revenue Ops AnalystQuerying your own data and turning the result into an answer.
Market Research DeskParallel fan-out across sources, then a synthesis step.
Incident Response TriageClassification and branching, where the branch decides what happens next.

There are more in the full template gallery on the Swarm Canvas — these four are simply the ones surfaced on the dashboard.

Workspace stats

Five tiles count what you own — Agents, Swarms, Chats, Tools, and Knowledge bases. Each tile links to the corresponding library page.

Spend & usage — by person, team or organisation

The Spend & usage panel attributes model cost, so it can be charged back rather than only totalled. Two pickers control it.

ScopeWho can pick itWhat it covers
Just meEveryoneYour own runs. The default.
My teamsAnyone in at least one IAM groupEveryone in the groups you belong to — resolved from your membership, not chosen by you.
Whole organisationSuperadmins onlyEvery user. Also the only scope that shows people outside your teams.

The picker offers only the scopes you may actually use. If you are in no team, “My teams” is absent rather than present-and-broken — and a scope you are not entitled to is refused, never quietly answered with your own numbers under someone else’s label.

The time range covers the last 24 hours through year to date. Windows are half-open and in UTC, so a run landing exactly on a boundary is counted once and the same dashboard reads the same from any timezone.

Team totals overlap on purpose

Someone in two teams contributes their spend to both, so the team rows do not add up to the total. That is the right answer to “what did this team cost” — the alternative is splitting one person's spend arbitrarily between teams, which is a worse lie than an overlap you can see.

Cost comes from the same column the budget caps read, so a figure here and a budget alert can never disagree about what someone spent.

Activity and model mix

The Activity panel charts hourly run volume across all your agents and swarms for the last 24 hours, with three summary figures underneath: success rate, average latency, and spend. The Model mix panel next to it breaks recent runs down by model, so you can see at a glance where your tokens went.

Recent runs

The last six executions across your workspace, each with the agent name, model, latency, cost, and a success/error indicator. View all opens the full run history at /traces, where every run can be expanded into its complete trace — see Logs & traces for how to read one.

A learning side panel ("Welcome — start here") rides along on the right of the dashboard with a short map of how the platform teaches agentic AI. Most authenticated screens have one of these panels, scoped to the screen you are on.

What to do on a new workspace

  1. 1

    Connect a model provider

    Integrations. Until you do, calls run on the operator's shared fallback key — fine for a first look, wrong for anything real. See Models & providers.
  2. 2

    Add data or documents

    Data Catalog for rows, Knowledge Base for prose. An agent with neither is just a chatbot.
  3. 3

    Build one agent

    Agent Builder — name, prompt, model, one or two tools.
  4. 4

    Run it and read the trace

    Logs & traces. This is the habit worth forming early.
  5. 5

    Before anyone else joins

    Turn off public signup and set budget caps — Access control and Budgets.