POC

Full cost analysis against the log analyzer before any deploy. The POC runs with read-only credentials and processes events locally — events stay on the machine.

"cost POC on our Splunk"

Pulling ~250K-event sample. Returned snapshot_id=snap_abc123. ETA 2–3 min.

"show report snap_abc123"

Top driver: Payment_Gateway_Timeout $4.2K/wk in payments-svc → recommend Reducer cap at 5%. Projected savings $306K/yr across the stack. Full report at /tmp/log10x-reports/poc_from_siem-snap_abc123.md.

You ask	Example answer
cost POC on our Splunk	Pulls a representative ~250K-event sample from your log analyzer (Splunk / Datadog / Elasticsearch / CloudWatch / Sumo / Azure / GCP / ClickHouse) in 2–3 min. Returns a snapshot_id.
show report snap_abc123	A 9-section report: top patterns by cost, cost per service, projected savings, per-pattern recommendations, suggested filter configs.
cost POC from kubectl, namespace production, 1h	Industry-pricing matrix built from a kubectl-sourced sample. Use when no log analyzer credentials are available yet. Sample-composition table calls out which pods dominate the bytes so the projection isn't trusted blind.

Prerequisites

Read-only credentials for your log analyzer for POC submit and POC status. Local POC needs kubectl on PATH instead. Privacy mode (default on every path) also needs the CLI installed locally (brew install log10x/tap/tenx).