A deliberate architecture for complete, cost-efficient log retention.
OBSESC runs entirely in your AWS account. It sits alongside your existing observability stack, receives a copy of your logs, and stores every event — using a compact navigation tier for speed and a complete fidelity tier for depth. No data leaves your account. No agents to deploy. No existing tools to replace.
A single deployment in your account. OBSESC integrates with your existing log shippers, stores data in your object storage, and works with your existing query engines.
No hosted control plane.
No access role for us.
One machine image,
sized to your volume.
A fan-out in your shipper.
Nothing removed.
Your data in open formats,
readable by your tools.
OBSESC separates speed from completeness. The navigation tier makes aggregate questions fast. The fidelity tier keeps every event, complete.
A single flow, from your existing telemetry to long-term answers.
A copy of your logs is sent from your existing shipper (OTel, Fluent Bit, Vector, etc.).
OBSESC builds a navigable map of your logs while keeping every original event.
The map goes to the navigation tier. Every event goes to your object storage.
Ask aggregate questions in seconds, or run SQL over every event when you need it.
Your data stays in your account, for as long as you need it.
