Runtime logs
Stream structured logs from every container as they happen, searchable and kept for as long as you need.
Observability
See everything your apps are doing
Runtime logs, distributed traces and metrics for every service you deploy, collected on infrastructure you own and kept in ClickHouse, so you can find out what just happened without shipping your data to someone else.
Three signals, one place
Logs, traces and metrics that belong together
Stop stitching a monitoring stack together by hand. Skitflow captures all three signals and keeps them side by side.
Distributed traces
Follow a request across services to see where the time goes and which hop broke.
Metrics
Track CPU, memory, latency and throughput per service, with the alerts you already rely on.
How it fits together
Agents collect, ClickHouse stores, you and the AI read
Telemetry agents gather every signal and ship it into a single store your dashboards and AI Assistant can both use.
Apps & nodes
Telemetry agentsOpenTelemetry
ClickHouseColumnar store
LogsTracesMetrics
Dashboards
AI Assistant
Agents run next to your workloads, so raw telemetry never leaves your servers. ClickHouse keeps it cheap to store and fast to query, even at long retention.
What you get
Built for the signals you actually chase
OpenTelemetry-native
Agents speak OpenTelemetry, so anything already instrumented for OTel just works.
ClickHouse-backed storage
A columnar store built for telemetry: fast queries, low cost, and retention measured in weeks, not hours.
Correlated signals
Jump from a slow trace to its logs and metrics without switching tools or losing context.
Self-hosted by default
Telemetry stays on infrastructure you control, and nothing is shipped off to a third-party SaaS.
Alerts & thresholds
Reuse the CPU, memory and custom thresholds from monitoring to get paged before your users notice.
AI-assisted root cause
The Assistant can use the same telemetry context to spot anomalies and suggest what to inspect first.
Observability questions
Do I need a separate monitoring stack?
No. Skitflow collects logs, traces and metrics itself, so you don't have to run and wire up Prometheus, Loki and Tempo by hand.
Why ClickHouse?
Telemetry is large and mostly write-once. ClickHouse's columnar engine keeps storage cheap and queries fast, which is why it is a common backend for logs and traces at scale.
What do the agents collect?
The telemetry agents gather runtime logs, distributed traces and metrics from your containers and nodes and forward them over OpenTelemetry.
Does this replace Monitoring and Swarm Monitoring?
It builds on them. Today's metrics and service map stay as they are; observability adds logs and traces and brings all three signals into one store.
Is my telemetry data private?
Yes. Agents run beside your workloads and write into your own ClickHouse, so your telemetry never leaves your infrastructure.
Know what your apps are doing
Bring logs, traces and metrics together on infrastructure you own, and let AI Assistant help you read them.