Splyntra
Integrations · OpenTelemetry Standards

Zero vendor lock-in with native OpenTelemetry OTLP

Don't trap your AI agent telemetry inside proprietary walled gardens. Splyntra is built from the ground up on standard OpenTelemetry GenAI semantic conventions. Ingest from any standard OTel exporter over gRPC or HTTP.

otel_env.sh
bash
# Point any standard OpenTelemetry exporter directly to Splyntra
export OTEL_EXPORTER_OTLP_ENDPOINT="https://ingest.splyntra.com"
export OTEL_EXPORTER_OTLP_HEADERS="x-splyntra-key=$SPLYNTRA_API_KEY"
export OTEL_SERVICE_NAME="customer-support-agent"

# Run your application in any language (Python, Node, Go, Rust, Java)
python main.py
100% OTel
Semantic Standards
Native GenAI OpenTelemetry specification.
gRPC + HTTP
OTLP Endpoints
High-throughput binary & JSON streams.
Zero Lock-In
Data Portability
Export telemetry to any OTel backend.
< 2ms
Collector Ingest
Ultra low latency ingestion pipeline.

Engineered for high-throughput autonomous agents

Every capability is built into the OpenTelemetry streaming pipeline with sub-millisecond ingestion overhead.

01Standards Compliant

GenAI Semantic Conventions Compliance

Full native support for OpenTelemetry's standardized GenAI attributes (`gen_ai.system`, `gen_ai.request.model`, `gen_ai.usage.prompt_tokens`, `gen_ai.usage.completion_tokens`).

  • Seamless compatibility with official OpenTelemetry SDKs
  • Automatic extraction of token usage, model parameters, and stop reasons
  • Zero translation overhead or custom proprietary wire formats
02OTLP Ingest

Direct OTLP gRPC & HTTP Collectors

Send traces, logs, and metrics directly to Splyntra's high-speed OTLP ingestion endpoints without needing proprietary client libraries.

  • High-throughput Protobuf gRPC streaming for production clusters
  • HTTP/JSON fallback endpoint for serverless and browser environments
  • Header-based organization and project routing (`x-splyntra-key`)
03Data Freedom

Bidirectional Data Pipeline & Export

Splyntra is not a data trap. Export your raw or enriched agent traces outward to ClickHouse, Datadog, Snowflake, S3, or Grafana Tempo.

  • Configure outbound OTLP exporters with custom filtering rules
  • Stream enriched risk scores and token costs to your enterprise SIEM
  • Maintain full ownership of your training and evaluation telemetry

Frequently Asked Questions

Can I use Splyntra if I already have an OpenTelemetry Collector deployed?
Yes. Simply add Splyntra as an OTLP exporter destination in your `otel-collector-config.yaml` pipeline.
What languages are supported?
Because Splyntra ingests standard OTLP, any language with an OpenTelemetry implementation (Python, TypeScript, Go, Java, Rust, C#, Ruby) works out of the box.

Ready to monitor and secure your AI agents?

Get started in under 3 minutes with zero credit card required. Free tier includes up to 5 projects and community telemetry.