Use case

Error monitoring that turns noise into issues

Report from any app, and Condux fingerprints events into deduplicated issues with real stack traces, breadcrumbs, tag facets and exact per-hour charts. Alerts fire the moment a new or regressed issue lands.

Get started free

The foundation everything else builds on. Send events with a Sentry SDK (change the DSN) or native OpenTelemetry logs, and Condux does the grouping, sampling and time-series so one bug is one issue with an exact count.

How it works

Ingest
A stateless relay authenticates, rate-limits and scrubs PII, then buffers through Redpanda.
Group
Events fingerprint by stack trace into issues; raw events are sampled, time-series stay exact.
Alert
New and regressed issues fire to email, Slack, Discord or a webhook, above a severity you set.

Report with or without an SDK

Sentry SDK (change the DSN) or first-party

# Any Sentry SDK: point it at Condux
SENTRY_DSN="https://<key>@ingest.condux.ai/<project>"

# Or no SDK at all: native OpenTelemetry logs
OTEL_EXPORTER_OTLP_ENDPOINT=https://ingest.condux.ai/api/<project>

Frequently asked questions

Are the event counts accurate?

Yes. Condux samples the raw event payloads it stores, but the per-hour time series counts every event, so your issue counts and charts are exact.

Keep exploring

AI-assisted bug fixing

Turn an issue into a fix PR.

Migrate from Sentry

Keep your SDK, change the DSN.

For developers

The developer-first tour.