Copilot AI Metrics — by Agile Analytics (Baytek)
Free · forecast your Copilot bill before June 1, 2026. GitHub Copilot usage reports, cost forecasting, seats, activity, chat, agent workflows, CLI, pull requests, local history, and an Action Center — all inside Azure DevOps. No backend. No data leaves your browser.
Ready for June 1, 2026?
Starting June 1, 2026 GitHub Copilot moves from request-based billing to usage-based billing. Every Copilot admin loses the cost predictability they have today. Copilot AI Metrics projects next-month spend from your live usage reports, flags users on track to overshoot, and tells you which seats to reclaim before the billing cycle closes — without a backend, a contract, or a credit card.
Overview
Copilot AI Metrics gives Azure DevOps admins a browser-only workspace for understanding GitHub Copilot adoption and spend without leaving ADO.
- Billing Readiness view: Forecast monthly cost under usage-based billing using your live activity, with per-user and per-team drill-down
- GitHub usage reports foundation: Uses the current Copilot usage reports model (GA Feb 2026), including team-level user mapping (May 2026), per-user CLI activity (Apr 2026), and active vs passive code reviewers (Apr 2026)
- Admin-first workspace: Built for org owners, engineering managers, and platform teams
- Frontend-only: Direct browser-to-GitHub API calls, no Baytek backend, no data relay
- Actionable reporting: Highlights seats to reclaim, users to coach, passive code reviewers to nudge, and wins to share
- Honest coverage: Calls out what is live, derived, estimated, or unavailable
Dashboard Areas
- Overview: KPI cards, executive readout, top insights, benchmark scoreboard, metric coverage
- Seats & Teams: Team adoption summary, seat status, inactivity, editor mix, cost visibility
- Trends: Time-series view for generated outputs, accepted outputs, active users, lines added, and chat activity
- AI Metrics / Overview: Org snapshot with live usage-report coverage and summary cards
- AI Metrics / User: User leaderboard, coaching candidates, engagement detail
- AI Metrics / Teams: Team comparison, adoption concentration, operational drill-down
- AI Metrics / Analysis: Language, model, PR, and workflow patterns
- AI Metrics / Actions: Action Center for reclaim, coach, watch, and celebrate follow-up
- AI Metrics / Methodology: Transparent formulas, caveats, and governance notes
- Settings: Demo/live switching, PAT validation, privacy notes, refresh preferences
Key Features
Current GitHub Copilot Usage Signals
- Licensed seats and active users
- Generated outputs, accepted outputs, and derived acceptance rate
- AI-assisted lines added
- Chat-style workflow usage
- Agent/edit workflow usage
- CLI usage
- Pull request activity, including Copilot-authored PR metrics when available
- Model and language breakdowns when present in source data
Action Center
- Reclaim seats: Spot never-activated or stale licenses
- Coach next: Find users or teams with low recent engagement
- Watch: Surface unusual drops or concentration risks
- Celebrate wins: Highlight strong adoption and workflow expansion
Each item includes the signal, evidence, and recommended next step.
Local History
- Stores local browser snapshots so you can compare recent states over time
- No external database required
- Useful for lightweight trend continuity beyond a single refresh
Demo and Live Modes
- Explore the full product with bundled demo data
- Connect your GitHub organization and PAT when you are ready
- Test connection before switching to live data
Export and Governance
- CSV export from the dashboard
- Explicit privacy and source notes
- Browser-only token storage
Privacy & Security
| What |
Detail |
| PAT storage |
Browser localStorage only |
| API calls |
Direct browser → api.github.com |
| Baytek relay |
None |
| Telemetry |
None |
| Reset |
Clearing settings removes local stored workspace data |
Requirements
| Requirement |
Detail |
| GitHub Copilot |
Business or Enterprise |
| GitHub PAT |
read:org or manage_billing:copilot |
| Azure DevOps |
Azure DevOps Services |
Quick Start
- Install the extension from the Marketplace.
- Open your Azure DevOps project and launch Copilot AI Metrics.
- Start with demo data or enter your GitHub org and PAT.
- Run Test Connection.
- Switch to live data and review the Action Center, seats, trends, and AI Metrics tabs.
Support
| |
