Open beta · Know your AI coding ROI

What changed

Release notes, new features, and fixes. Shipped in order, without the noise.

Know if your AI coding spend is paying off

Backlight is open. Connect your tools, sign up, and see what your AI coding spend returns — in cost, in output, and in code that holds up.

You see cost per AI engineer with a per-vendor breakdown, license utilization, idle seats, and how spend tracks against output over time. Then you decide what to keep paying for.

  • Is the spend working. Cost per AI engineer by vendor, license utilization, idle-seat detection, and an AI spend-versus-output trend.
  • AI versus human, honestly. A free heuristic classifier reads commit metadata to separate AI-assisted work from human-only work. Team adds optional LLM classification for higher precision.
  • Delivery and code health. DORA-style delivery metrics including lead time, change failure rate, and time to restore. A code health score, regression rate, and output per engineer. Five dashboards: Overview, Productivity, Code Health, Financial Operations, and Organization.
  • Your stack, connected. GitHub, GitLab, Bitbucket, Azure DevOps, Jira, Linear, and Slack. Connect your AI coding tools — Copilot, Claude Code, ChatGPT, and Cursor — to see usage and spend. 2FA and GDPR data export for everyone; SSO/SAML and an audit log on Enterprise.
  • Free. Analytics for small teams: 90-day history, up to 10 active engineers, every integration, full AI ROI, and the free heuristic classifier.
  • Team. Adds LLM AI and bug classification, 365-day history, API access, CSV exports, and configurable alerts.
  • Enterprise. SSO/SAML, audit log, advanced retention, and unlimited history. Talk to sales.

Self-serve sign-up is open today. See plans and pricing for the full breakdown.