Rationale

1. It provides a granular (per-GB) consumption pricing model, which scales perfectly on horizontal-based architectures like ours.
   
2. It supports OpenTelemetry, allowing us to keep observability standardized.
3. It supports Kubernetes, which further increases our infrastructure observability.
   
4. It supports logs, allowing us to keep a centralized place for execution logs, improving debugging experience.
   
5. It records users' sessions, which comes in handy when replicating bugs and researching errors.

Alternatives

1. Datadog: It provided a per-host pricing model, this made it much more expensive due to our horizontal approach towards scaling.
   
2. Atatus: It provided a per-host pricing model, this made it much more expensive due to our horizontal approach towards scaling. It did not support Kubernetes, reducing infrastructure observability. It did not support logs, decreasing debugging experience. Code traces were not always collected, reducing debugging experience.
   
3. New Relic: Code traces were not always collected, reducing debugging experience.

Usage

1. Platform backend.
   
2. Scanner executions.
   
3. Kubernetes infrastructure.
   
4. Real User Monitoring (RUM): web vitals and session recordings for our website and platform.