For Org

This view is available by switching to Team (top right, next to "Me") in Session Insights. It provides five tabs, each addressing a different question about the organization as a whole, rather than an individual's own work.

Briefing

An overview of the current window's total spend and its source, a list of gaps requiring attention (people who have never reported, an unmatched identity), adoption by team, and the "movers": the people whose spend increased the most. Selecting a mover's row opens their full record: tokens, streak, most-used model, and a year of activity.

Adoption & Impact

Coverage data showing who on the roster is reporting, broken down by team, with a per-developer table listing the signal each person reports through, along with cost, sessions, and machines. This view answers whether the rollout is actually being used, rather than only whether it has been enabled. Below the roster, delivery-efficiency metrics tied to git are shown, including cost per commit, per pull request, and per 1,000 lines changed, the acceptance rate of AI-suggested edits, and the same figures broken down by branch.

Economics

A breakdown of spend: the total for the selected window, a daily chart split by team, per-head averages, and a ranked list of cost by person (selecting a row opens that person's full record). Below this, spend is broken down by team or by model, including cache hit rate, cost, cost per member, and share of total spend, along with budgets, model policy, and the signal each figure is sourced from.

Teams

Team membership and working patterns, presented in three views:

• Roster: Create teams and assign members manually, or import proposed assignments from machine tags or SCIM groups. A manual assignment always takes precedence over an imported one.
• Practice: Each team's scored dimensions (steering, delivery, planning, quality, design), along with coverage and the number of people the score is based on.
• Benchmark: Up to four teams compared against each other and against the workspace as a whole, shown as a chart and a ranked table.

Data & rollout

A four-step process for enabling Session Insights across the organization:

1. Sources: Shows the status of the two signals that feed all of the data above: the session export (run per developer via the CLI) and live OpenTelemetry data (submitted automatically once configured, with no separate installation required).
2. Teams: Links to the Roster view described above; there is nothing to configure directly on this step.
3. Governance: Configures who can view a named individual's scores (administrators only, administrators and leads, or nobody), the minimum group size, whether leads are restricted to their own team, how the system handles a person reporting through both signals at once, an optional model allowlist, and a weekly digest.
4. Rollout: Provides the configuration to hand to whoever manages device deployment (Jamf, Intune, or similar), with one tab per tool: Claude Code, Cursor Enterprise, Codex CLI, and Gemini CLI. Each file-based tab provides a ready-to-copy file specifying the enforcement level and identity rule. Cursor Enterprise is set up in Cursor's Team Settings under OpenTelemetry Export instead of through a file, so this configuration does not need to be reverse-engineered from the product itself. Below the templates is a plain-language explanation of what data is collected and retained, along with a tool for verifying a usage export's signature.

Next