Technology Overviews
Supervising coding agents with Herdr
Check agent status, terminal persistence, and recovery before adopting Herdr.
Practical guidance for SRE and platform teams operating reliable services and production AI. Independent analysis, field guides, and working resources.
Selected reading
Evaluate incident-response agents against service outcomes
Build a small evaluation that can reject a convincing but unverified recovery claim, with independent service checks and a runnable Python grader.
A practical starting point for evaluating agents before expanding their operational role.
Read the article
Original writing / Newest first
More from AIOpsSRE
AIOps
What MTTD misses
Separate detection from response and find delays that an improving average can conceal.
AIOps
When agents can change production
Define the action, contain its scope, and verify what changed at the destination.
How-To
Connect the signals with OpenTelemetry
Use instrumentation and context propagation to connect the evidence in an investigation.
How-To
The SRE and platform handoff
Clarify ownership across shared platforms, service reliability, and incident response.
How-To
What did the postmortem fix prove?
Separate a completed task from evidence that the underlying risk was reduced.
Releases, research & operational impact
Latest AI news
AI inferenceNews analysis
NVIDIA’s confidential inference results need a workload boundary
NVIDIA reports low confidential-computing overhead on an eight-GPU test. Separate that measured performance result from your own security and recovery claims.
Read the analysis
AI infrastructureNews analysis
NVIDIA Topograph puts GPU placement evidence in focus
New Topograph guidance connects physical network topology to workload schedulers. The operational check is whether the scheduler’s view survives cluster change.
AI model operationsNews analysis
GPT-6 Sol and Luna change the agent cost calculation
Lower model prices and new caching controls make agent pilots cheaper to run. Measure accepted work before expanding the workload.
From reading to doing
Work through it, step by step.
Incident response
Build a usable runbook
Work through a database connection example, from the first symptom to a verified recovery.
Worked example + Markdown templateRead the guide
Reliability planning
Put your error budget to work
Define the service objective, calculate the allowance, and agree when a release needs to wait.
Calculations + policy fieldsRead the guide
Observability
Trace a request across services
Start with one important request path. Check propagation, sampling, and the gaps in your evidence.
Implementation guideRead the guide