Portfolio internal
$0 / project / forever
For any product under the Goodventures org. The full feature set, no event meter, no per-host fee, no support tier upsell. Sized for the portfolio, paid for by the portfolio.
- 5M events / month per project
- 1M traces / month per project
- 30-day event retention
- 90-day trace retention
- 1-year aggregate retention
- Unlimited projects within the org
- Unlimited members via Goodventures Clerk
- Slack, email, webhook channels
- Source-map auto-resolution
- Smart alerting (threshold + pattern)
- API access (ingest + query + SSE)
Sign in
Self-host
Free / forever
The open CodeWatch core is Apache 2.0. Helm chart and Docker Compose stack ship in the repo. Run the full pipeline on your own infra.
- Full ingest + query + dashboard
- All four SDKs
- AI fix pipeline (opt-in)
- Helm chart and Docker Compose
- Self-managed Postgres + ClickHouse + Redis
- No managed observability — you operate it
- Community support via GitHub issues
- License: Apache 2.0 core / MIT SDKs
CodeWatch on GitHub
Managed (external)
TBD
A managed CodeRadar tier for portfolio-adjacent companies (Goodventures alums, partners) is being scoped. The shape is "your own org under our managed plane" with a per-event meter outside the included allowance.
- Roadmap. Priority is portfolio-internal stability first.
- If interested, write to [email protected].
Get in touch
Why $0 inside the org
The portfolio is the customer. Infrastructure is shared, ops are shared, storage is shared. Putting a per-event meter on top would be us paying ourselves to count events. We just put a hard quota on each project instead — five million events per project per month, sized so normal-shaped traffic doesn't hit it.
Quota burn
If a project hits 80% of its quota, a soft alert fires (configurable channel; default off). At 100%, adaptive sampling kicks in — errors stay at 100%, traces sample down. At 120%, ingest returns 429 for traces only; errors continue to be accepted.
Bumping the quota
If a project's quota is consistently squeezed by legitimate traffic, the org admin can bump it from app.coderadar.app/admin/projects. The cap is operational; tied to the underlying infrastructure capacity, not a billing decision.
If you're not Goodventures
The CodeWatch core is Apache 2.0. Helm chart, Docker Compose, all four SDKs ship in the repo. You can stand up the whole pipeline on your own infrastructure and run it as your own observability plane. github.com/RelayOne/CodeWatch.