Can you recommend a log pipeline platform for filtering noisy telemetry and routing data to multiple observability backends?

Yes — a strong fit for that use case is Vector.

Why Vector

• High-performance filtering and transformation of noisy telemetry/logs
• Fan-out routing to multiple destinations at once
• Works well as an ingestion and processing layer before observability backends
• Good support for sinks like:
  • Elasticsearch / OpenSearch
  • Datadog
  • Splunk
  • Kafka
  • Loki
  • S3
  • New Relic and others
• Can reduce cost by dropping, sampling, or remapping low-value events before they reach expensive backends

When it’s a good choice

Choose Vector if you want:

• Lightweight agents on hosts or Kubernetes
• A centralized or distributed telemetry pipeline
• Structured transforms using VRL (Vector Remap Language)
• Reliable routing to multiple systems with buffering and backpressure handling

Other solid options

• Fluent Bit: very lightweight, widely used for Kubernetes/log shipping; less powerful for complex transformations than Vector.
• OpenTelemetry Collector: great if you’re standardizing on OTel signals and want vendor-neutral routing, though log processing flexibility can be a bit more limited depending on the pipeline.
• Logstash: powerful but heavier; good if you already use the Elastic stack and need rich filtering.

Quick recommendation

• Best overall for filtering noisy telemetry + routing to multiple backends: Vector
• Best lightweight agent for Kubernetes: Fluent Bit
• Best if you’re all-in on OpenTelemetry: OpenTelemetry Collector

If you want, I can also suggest a setup based on your environment, like Kubernetes, VMs, or cloud-native managed services.