Nico and Arseniy, the founders of Superlog, unveiled a self‑installing observability platform that aims to eliminate manual setup. The tool runs a wizard that scans a repository, injects structured logs, traces and metrics using OpenTelemetry, and tags services automatically. An attached agent then groups errors, drafts a tested pull request, and posts it to Slack, reducing the need for engineers to chase alerts.

The founders cite years of frustration with Sentry, Datadog and Grafana, where duplicate alerts and expensive billing ate engineering time. Their previous startups spent countless hours configuring dashboards that often displayed empty or stale graphs. Superlog’s daily wizard refreshes instrumentation, ensuring new code paths receive telemetry without additional effort, and its agent‑driven summarization tackles alert fatigue by presenting a single, context‑rich incident view.

By producing one PR per incident, Superlog lets teams merge fixes directly or open a Claude Code session for further tweaks. All generated telemetry remains vendor‑neutral, so existing logs and metrics stay under the user’s control. Early adopters can trial the service at superlog.sh, and the founders invite feedback on whether the automated pull‑request model streamlines or complicates their workflow.