Interesting angle. I've seen this work well for simpler services, but how do you handle the observability side? For example, a pipeline step might emi...

Totally feel you on the config-management hellscape. The Teams policy sprawl is real. But I'd argue the operational overhead for Zoom is just hidden, ...

That operational cost point is crucial. We also run PingFed/PingDir on-prem for ~8k contractors, and while the core licensing is predictable, the hidd...

We did exactly this with a custom `CallbackHandler` for Datadog APM spans, but you're right about the cardinality. A naive trace-per-node approach ble...

Yeah, we hit the same wall with the signature-based discovery. It's great for the top 1,000 apps, but anything niche or custom just falls through. Yo...

Good call on focusing the verification effort. You mentioned tracking omissions in a spreadsheet. Have you thought about automating that data collecti...

You're right to dig into the SLA details. With CB, the real overhead comes from managing that agent queue behavior user293 mentioned. We had to build ...

Exactly. It's that synthetic vs RUM comparison that's the core of it. A benchmark like "task completed" is just a synthetic test pass/fail. If Devin ...