Real teams, real telemetry
Stories from our customer work, told with the numbers: what we found in their OpenTelemetry® data, what changed at the source, and what is still left to do. Every story is anonymized and every story is real.
Enterprise software · North America · Anonymized, but real · Audio episode
Nearly half the spans came from a service called cli: what we found at a 500-service enterprise platform
One background worker named cli made 45% of all spans, a third of logs were exact duplicates, a third were access logs, and the Collector shipped its own logs again.
- 45%of all spans came from one unnamed worker
- 32%of log records were exact duplicates
- 4.5%of logs were the Collector re-shipping its own output
Digital health · Europe · Anonymized, but real · Audio episode
The PII leak was a logger setting, not a developer problem: what we found at a European digital-health platform
Personal data in 1 in 6 services traced to one ORM logging setting, logs with no severity at all, one worker writing most of the logs, and auth tokens Rose found in a day.
- 17%of services flagged for personal data in week one
- 100klog lines a week with personal data, from one service
- 100%of logs arrived with no severity
Enterprise software · Latin America · Anonymized, but real · Audio episode
How a B2B SaaS platform cut logs 24% and metrics 25% in six weeks, without touching a single customer log
Repeated logs, metrics that never change, and a pipeline complaining about itself. How replay-verified Collector changes beat a six-week proof of value, and what happened after.
- 24%fewer log records
- 25%fewer metric datapoints
- 61%less volume on dedup-eligible logs
How we write these stories
- Every story comes from a real OllyGarden engagement. We remove the customer's name, people, products, and anything else that could identify them, and round numbers that could.
- Numbers come from the telemetry samples and measurements we took during the work, reported as ratios. We say how each one was measured and what it does not claim.
- We include what did not go well and what is still open, not only the wins.
The findings in these stories fall into five dimensions of telemetry quality: identity, volume and waste, cost drivers, safety, and correctness. What is telemetry quality? →
Want a story like thisfor your telemetry?
Send us a sample and we will show you where your volume, cost, and risk really come from.