Zhivko Todorov
ALL CASE STUDIES

CASE 91 · HAWTHORN · 2024

METRIC STREAMSCLOUDWATCHOBSERVABILITYORG

CloudWatch metrics, streamed centrally, queryable everywhere.

A B2B platform had observability per account — each team kept its own CloudWatch dashboards in its own account. Cross-account incident correlation took an engineer half a day per incident. We turned on CloudWatch Metric Streams across the org and landed everything in a central Prometheus-compatible store.

INDUSTRY

B2B platform

DOMAIN

LANDING ZONE

DELIVERED

2024

STACK

CLOUDWATCH METRIC STREAMS·KINESIS FIREHOSE·AMAZON MANAGED GRAFANA·AMP·S3

RESULTS

What changed, by the numbers.

CROSS-ACCOUNT CORRELATION

< 2m

WAS HALF A DAY

ACCOUNTS STREAMING

29

AT ROLLOUT

INCIDENT MTTR

−38%

CROSS-ACCOUNT INCIDENTS

METRIC RETENTION

13 MONTHS

AMP DEFAULT

HOW IT WENT

Cross-account incidents were the painful ones. The customer-facing API in account A was timing out; the database in account B was slow; the queue in account C was backed up. Three accounts, three CloudWatch consoles, three engineers swapping screenshots in Slack.

Metric Streams pushed every account’s metrics to a central Firehose, with Prometheus-format conversion at the Kinesis Data Stream consumer. Managed Grafana on top sat in one place; engineers built dashboards that spanned accounts as easily as services.

Cross-account correlation time dropped from half a day to under two minutes. Incident MTTR on cross-account incidents fell 38%. The "swap screenshots in Slack" pattern got replaced with "share this Grafana link."

READY WHEN YOU ARE

Let's get your AWS bill (and architecture) in order.

The discovery call is free. You walk away with at least one concrete idea — even if we never work together.

Or email directly →