Superlog is an AI-native observability platform built for engineering teams that want signal without the babysitting. A wizard scans the repository and installs production-grade OpenTelemetry instrumentation with proper semantic conventions, service tagging, and environment metadata. The system then runs daily check-ins that keep adding logs, alerts, and dashboards as the code evolves, so observability does not rot the moment a sprint ends.
When something breaks, Superlog groups related errors into a single incident, investigates with full context drawn from logs, traces, recent deployments, and historical Slack threads, then posts a mergeable pull request directly into Slack. Engineers review and merge instead of paging on call and digging through dashboards at 3 a.m. Telemetry remains vendor-neutral, so teams keep every log, trace, and metric even if they decide to leave the platform.
Founded in 2026 by Nicolo Magnante and Arseniy Shishaev, a former Datadog engineer who lived the on-call pain personally, Superlog is backed by Y Combinator in the Spring 2026 P26 batch. The product targets startups and scale-ups that ship fast, run on Kubernetes or serverless, and cannot justify a dedicated observability team. It positions itself as the autonomous layer between raw OpenTelemetry pipelines and incident response, replacing both the toil of instrumenting code and the toil of triaging alerts.