🪢 Open source agent evals & observability: Trace, evaluate, and improve LLM applications with one open platform.
TypeScript analyticsautogenevaluationlangchainlarge-language-modelsllama-indexllmllm-evaluationllm-observabilityllmopsmonitoringobservabilityopen-sourceopenaiplaygroundprompt-engineeringprompt-managementself-hostedycombinator
/* eslint-disable @typescript-eslint/no-use-before-define */
import { upsertDefaultModelPrices } from "./scripts/upsertDefaultModelPrices";
import { upsertLangfuseDashboards } from "./scripts/upsertLangfuseDashboards";
import { hello, initTelemetry } from "@langfuse/native";
import {
initializeClickhouseCompatibility,
logger,
recordIncrement,
} from "@langfuse/shared/src/server";
export const initializeWorker = async (): Promise<void> => {
initializeNativeAddon();
await initializeClickhouseCompatibility();
await Promise.all([upsertDefaultModelPrices(), upsertLangfuseDashboards()]);
};
// Two separate success signals. The addon must load and run, or the build is
// broken: hello() is deliberately unguarded and stops the worker. Whether the
// addon's own telemetry could be set up is an operational concern that must
// not keep the worker from processing jobs, so that failure is reported
// through the Node side's logger and metrics instead (the addon records its
// own metrics and logs once initialised; see packages/native/README.md).
const initializeNativeAddon = (): void => {
try {
initTelemetry();
} catch (error) {
logger.error("Native telemetry initialisation failed", {
error: error instanceof Error ? error.message : String(error),
});
recordIncrement("langfuse.native.telemetry_init_failed");
}
hello("startup");
};