Start free. Scale when it earns its place.

Throughline derives all four metrics from events your tools already emit — no agents, no manual logging. Cycle time is the wall-clock duration from first commit on a branch to deploy. Lead time for changes is the duration from first commit to when the change reaches production. Deploy frequency counts successful production deploys per day across all connected pipelines. PR throughput counts merged pull requests per week per team.

Stage breakdowns (coding, review wait, CI and checks, deploy to production) are derived by timestamping each state transition: branch creation, PR open, first review, last approval, merge, and deploy event. The p50 and p95 for each stage are computed over the selected window — defaulting to the trailing 30 days — and benchmarked against your own 90-day baseline.

Throughline connects via read-only OAuth scopes — it never writes to your repositories, opens pull requests, or modifies any code. For GitHub, it requests repo:read, pull_requests:read, and actions:read. For GitLab, the equivalent read-only project and pipeline scopes.

No SSH keys, no deploy keys, and no webhook secrets are stored. Connection takes under five minutes: authorize the OAuth app, select the repos you want to track, and your first metrics appear within minutes as Throughline backfills available history. SOC 2 Type II documentation is available on request.

A seat is any user who can log in to your Throughline workspace and view data. Read-only share links — public URLs you can paste into a doc or Slack — do not consume a seat. The Free plan allows up to 5 seats. Team and Enterprise plans have no seat limit.

Service accounts and bot users used only for the OAuth integration are never counted as seats. You can also create read-only member roles for stakeholders who need to view dashboards but should not change settings or connect new repos.

Yes — self-hosting is available on the Enterprise plan. You deploy Throughline inside your own VPC (AWS, GCP, or Azure) or on-prem, and all Git and CI data stays within your network boundary. We provide a Docker Compose reference configuration and a Helm chart for Kubernetes clusters, plus a dedicated Slack channel during initial rollout.

Self-hosted instances receive the same version cadence as the cloud product. Security patches are delivered as container image updates; your team controls the upgrade schedule. Data residency guarantees — US or EU — are available whether you are on cloud or self-hosted.

Yes. You can upgrade from Free to Team immediately from your workspace settings — the change takes effect the same day and you are billed a prorated amount for the remainder of your billing cycle. Downgrading from Team to Free takes effect at the end of your current billing period; history beyond 14 days is retained for 30 days in case you change your mind.

There are no long-term commitments on monthly billing. Annual billing carries a ~2-month discount and renews automatically unless cancelled at least 14 days before the renewal date. Enterprise contracts are negotiated separately and can include custom terms, purchase orders, and invoicing.

If you downgrade from Team to Free, your workspace immediately reverts to a 14-day rolling window. Historical data beyond 14 days is not deleted — it is retained for 30 days while you decide whether to reactivate. After 30 days, data older than 14 days is permanently deleted from our systems.

If you cancel your account entirely, all data is deleted within 30 days of cancellation. You can request an export of your metrics in JSON or CSV format at any time before deletion. Enterprise customers can negotiate custom data retention and deletion terms in their contract.