Direct OTLP ingest¶

Skip the third-party platform entirely: point your app's OpenTelemetry exporter at GigaFlow's OTLP receiver. No datasource/sync — spans arrive in real time.

Endpoint¶

HTTP: POST https://api.gigaflow.io/v1/traces (accepts application/json and application/x-protobuf). Logs: POST .../v1/logs.
gRPC (:4317) runs locally / under docker-compose but is not exposed on the hosted endpoint yet — use HTTP in production.

1. Create a project + mint an OTLP token¶

PID=$(curl -s -X POST "$GIGAFLOW_BACKEND_URL/projects/" -H "Authorization: Bearer $GIGAFLOW_API_KEY" \
  -H 'Content-Type: application/json' -d '{"name":"my-otlp-project"}' | python3 -c 'import sys,json;print(json.load(sys.stdin)["project_id"])')

# Mint a per-project OTLP token (returned ONCE, prefixed gflw_otlp_):
curl -s -X POST "$GIGAFLOW_BACKEND_URL/projects/$PID/otlp_tokens" \
  -H "Authorization: Bearer $GIGAFLOW_API_KEY"

The token scopes ingest to this project (clients can't smuggle traces across projects). Manage with GET/DELETE .../projects/$PID/otlp_tokens.

2. Point your exporter at GigaFlow¶

export OTEL_EXPORTER_OTLP_ENDPOINT=https://api.gigaflow.io
export OTEL_EXPORTER_OTLP_HEADERS="Authorization=Bearer gflw_otlp_..."
export OTEL_EXPORTER_OTLP_PROTOCOL=http/protobuf

(Use the OpenInference / OpenLLMetry / pydantic-ai instrumentation for your stack.)

3. Transform + compute¶

Set the project's transform to match your exporter's span shape (PUT /projects/$PID/transform). Then traces appear without sync:

gigaflow compute "SELECT trace_id FROM trace_metrics WHERE run_id IS NULL"
gigaflow inspect <trace_id>

See the backend's docs/otlp-quickstart.md for exporter snippets per framework.