Skip to main content

Agent Studio

Agent Studio is your monitoring and analytics dashboard for AI agent usage across your workspace.

Overview

Agent Studio gives you visibility into how your team uses AI across notebooks and conversations. Track usage, monitor costs, and review conversation history — all from a single dashboard.

Key Metrics

Four headline cards summarize your workspace's AI activity:

MetricDescription
ConversationsTotal AI conversations across all agent types
Total TokensAggregated token usage
Active UsersUnique team members using the AI agent
Cache Hit RatePercentage of cached responses, reducing costs

Each card includes a sparkline showing the trend over your selected time period.

Filtering

Filter the dashboard by:

  • Agent type — All, Notebook, Semantic, or Conversation
  • User — View activity for a specific team member
  • Time period — 7 days, 30 days, 90 days, or all time

Filters are persisted in the URL, so you can bookmark or share specific views.

Usage Charts

Conversations Over Time

A stacked bar chart showing conversation volume by agent type. The chart automatically switches between hourly and daily resolution based on your selected time range.

Top Users

A horizontal bar chart showing the most active users by conversation count.

Token Usage by Model

A grouped bar chart breaking down token usage per model:

  • Input (uncached) — Full-cost input tokens
  • Input (cached) — Cached tokens that save 90% on input costs
  • Output — Generated response tokens

Conversation History

A paginated table of recent conversations with:

  • Timestamp
  • User
  • Agent type (Notebook, Semantic, or Conversation)
  • First message preview
  • Message count

Click any row to open the full conversation, including all messages with timestamps and role indicators. From the detail view, you can jump directly to the source notebook or conversation.

Use Cases

Cost Optimization

Monitor token usage and cache hit rates to understand your AI spending. High cache rates mean the agent is efficiently reusing prior context.

Team Adoption

Track active users and conversation volume to understand how your team adopts AI-assisted analysis.

Quality Review

Browse conversation history to review the questions your team asks and the answers they receive. This helps identify training opportunities and common patterns.