The telemetry was there.
Nobody was watching.
You deploy on Friday, everything's green, you close the laptop. Saturday morning your phone buzzes — something has been broken for twelve hours and nobody noticed.
Not because the data wasn't there. Your logs, metrics and traces were sitting in some tool a teammate set up six months ago. The signals were flowing. No one — and nothing — was watching them.
That's the gap Epok exists to fill.
The observability industry has a fixation on storage and search — bigger indexes, faster queries, more dashboards. But the hard question was never “can I find this log line?” It's two other questions: did something just break? and, when it did, why?
Broad observability platforms can answer those questions, but teams often need to choose and configure the right instrumentation, monitors, queries, and workflows. Epok makes the incident operating loop available without requiring a dedicated observability program: start with the signals you have, then add depth as it proves useful.
Telemetry and incident analysis together.
Collect and explore your telemetry, detect changes, and investigate the evidence behind an incident.
Keep the evidence together
Store and search logs, metrics, traces and RUM. Explore the signals behind an incident in the same backend.
Catch changes automatically
New-error evidence, a service gone silent, a volume anomaly, or a latency break. Rule packs and learned detectors activate according to the signals and history available.
Inspect the probable cause
Epok correlates symptoms and ranks probable causes with deterministic scoring. Review the evidence and uncertainty alongside the AI-written explanation when available.
Epok is a complete backend for logs, metrics, traces and RUM, with incident analysis built in. It stores your telemetry and includes search, dashboards and SLOs. During evaluation, dual-ship to Epok beside your current tool. Move fully when it has earned your trust.
Signals arrive over every protocol that matters — OTLP, Loki, Elasticsearch bulk, FluentBit, Fluentd, syslog, CloudWatch, Prometheus remote-write, raw JSON. If you can send HTTP, you can send to Epok. No proprietary agents, no lock-in.
Observe, don’t wait
Immediate rule packs can match arriving evidence; learned detectors sharpen as they learn your normal. Threshold rules remain available for hard business constraints.
Day-one value
Begin with an approved telemetry boundary and no dashboard migration. The first useful result should come from the incident output, not from rebuilding your existing estate.
Root cause, not just alerts
Anyone can tell you something is red. Epok tells you which change, which service, and why — with every conclusion linked to the evidence behind it.
Honest by discipline
Every detector and suppression rule traces back to a specific miss or false page we found on our own always-on test fleet. Fewer false pages, earned the hard way.
Predictable cost
Plans start at $199/month with included unified volume and published $0.20/GB paid overage. No per-host, per-query, or cardinality line.
Speed is a feature
At 2am, the gap between fixing before the SLO breaches and after is how fast you can search, correlate, and act.
We built Epok because we needed it ourselves — and because the tools that could do this were priced and staffed for companies ten times our size.
The engine has run for months against our own always-on test fleet and injected faults. Every detector and every suppression rule exists because something specific went wrong there, and the next version was built to catch it — or to stop it from crying wolf. That's the whole product: telling you something is wrong, and why, without you asking, and without paging you for things that aren't.
Reach the team: [email protected] — answers within a day, no sales filter.
Why We Built Epok
The longer version of the story — what we tried, what failed, and what we decided to do about it.
Catch New Errors Before Users Report Them
How automatic error fingerprinting works, and why it beats error counting.
Silent Failures: The Bug That Won’t Page You
Why absence is the most dangerous signal in production, and how to detect it.
We built Epok because we needed it. We think you might too.