Class LdAiConfigTracker

Records metrics for a single AI run.

Inheritance

object

LdAiConfigTracker

Implements

ILdAiConfigTracker

Inherited Members

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

object.ReferenceEquals(object, object)

object.ToString()

Namespace: LaunchDarkly.Sdk.Server.Ai

Assembly: LaunchDarkly.ServerSdk.Ai.dll

Syntax

public class LdAiConfigTracker : ILdAiConfigTracker

Remarks

All events a tracker emits share a runId (a UUIDv4) so LaunchDarkly can correlate them in metrics views. See individual track methods for their specific semantics. Call CreateTracker on the AI Config to start a new run. A ResumptionToken preserves the runId, so events emitted by a tracker reconstructed in another process correlate with the original tracker's runId.

Properties

| Edit this page View Source

ResumptionToken

A URL-safe Base64-encoded token that can be used to reconstruct this tracker in a different process or at a later time. The token contains the runId, configKey, variationKey, and version.

Use LdAiConfigTracker.FromResumptionToken to reconstruct a tracker from this token.

Declaration

public string ResumptionToken { get; }

Property Value

Type	Description
string

| Edit this page View Source

Summary

A summary of the metrics tracked by this tracker.

Declaration

public MetricSummary Summary { get; }

Property Value

Type	Description
MetricSummary

Methods

| Edit this page View Source

FromResumptionToken(string, ILaunchDarklyClient, Context)

Reconstructs a tracker from a resumption token. This enables cross-process scenarios such as deferred feedback, where a tracker's runId needs to be reused in a different process or at a later time.

The reconstructed tracker will have empty model and provider names, as these are not included in the resumption token.

Declaration

public static LdAiConfigTracker FromResumptionToken(string token, ILaunchDarklyClient client, Context context)

Parameters

Type	Name	Description
string	token	the resumption token obtained from ResumptionToken
ILaunchDarklyClient	client	the LaunchDarkly client
Context	context	the context to use for track events

Returns

Type	Description
LdAiConfigTracker	a new tracker associated with the original run

Exceptions

Type	Condition
ArgumentNullException	thrown if token or client is null
ArgumentException	thrown if the token is malformed or missing required fields

| Edit this page View Source

TrackDuration(float)

Tracks a duration metric related to this config. For example, if a particular operation related to usage of the AI model takes 100ms, this can be tracked and made available in LaunchDarkly.

Declaration

public void TrackDuration(float durationMs)

Parameters

Type	Name	Description
float	durationMs	the duration in milliseconds

Remarks

Records at most once per Tracker; further calls are ignored.

| Edit this page View Source

TrackDurationOfTask<T>(Task<T>)

Tracks the duration of a task, and returns the result of the task.

If the provided task throws, then this method will also throw.

In the case the provided function throws, this function will still record the duration.

Declaration

public Task<T> TrackDurationOfTask<T>(Task<T> task)

Parameters

Type	Name	Description
Task<T>	task	the task

Returns

Type	Description
Task<T>	the task

Type Parameters

Name	Description
T	type of the task's result

| Edit this page View Source

TrackError()

Tracks an unsuccessful generation event related to this config.

Declaration

public void TrackError()

Remarks

Records at most once per Tracker. TrackSuccess and TrackError share state; only one of the two can record per Tracker, and subsequent calls are ignored.

| Edit this page View Source

TrackFeedback(Feedback)

Tracks feedback (positive or negative) related to the output of the model.

Declaration

public void TrackFeedback(Feedback feedback)

Parameters

Type	Name	Description
Feedback	feedback	the feedback

Remarks

Records at most once per Tracker; further calls are ignored.

Exceptions

Type	Condition
ArgumentOutOfRangeException	thrown if the feedback value is not Positive or Negative

| Edit this page View Source

TrackRequest(Task<Response>)

Tracks a request to a provider. The request is a task that returns a Response, which contains information about the request such as token usage and metrics.

This function will track the duration of the operation, the token usage, and the success or error status.

If the provided function throws, then this method will also throw.

In the case the provided function throws, this function will record the duration and an error.

A failed operation will not have any token usage data.

It is the responsibility of the caller to fill in the Response object with any details that should be tracked.

Example:

var response = tracker.TrackRequest(Task.Run(() => new Response {
    // 1. Make a request to the AI provider
    // 2. Identify relevant statistics returned in the response
    // 3. Return a Response object containing the relevant statistics
    Usage = new Usage { Total = 1, Input = 1, Output = 1 },
    Metrics = new Metrics { LatencyMs = 100 }
}));

If no latency statistic is explicitly returned in the Response, then the duration of the Task is automatically measured and recorded as the latency metric associated with this request.

Declaration

public Task<Response> TrackRequest(Task<Response> request)

Parameters

Type	Name	Description
Task<Response>	request	a task representing the request

Returns

Type	Description
Task<Response>	the task

Remarks

Subsequent calls re-run the task but emit only metrics not already recorded on this Tracker. Call CreateTracker on the AI Config to start a new run.

| Edit this page View Source

TrackSuccess()

Tracks a generation event related to this config.

Declaration

public void TrackSuccess()

Remarks

Records at most once per Tracker. TrackSuccess and TrackError share state; only one of the two can record per Tracker, and subsequent calls are ignored.

| Edit this page View Source

TrackTimeToFirstToken(float)

Tracks the time it takes for the first token to be generated.

Declaration

public void TrackTimeToFirstToken(float timeToFirstTokenMs)

Parameters

Type	Name	Description
float	timeToFirstTokenMs	the duration in milliseconds

Remarks

Records at most once per Tracker; further calls are ignored.

| Edit this page View Source

TrackTokens(Usage)

Tracks token usage related to this config.

Declaration

public void TrackTokens(Usage usage)

Parameters

Type	Name	Description
Usage	usage	the token usage

Remarks

Records at most once per Tracker; further calls are ignored.

Implements

ILdAiConfigTracker

Extension Methods

PluginExtensions.GetPluginHooks<TClient, THook>(TClient, IEnumerable<PluginBase<TClient, THook>>, EnvironmentMetadata, Logger)

PluginExtensions.RegisterPlugins<TClient, THook>(TClient, IEnumerable<PluginBase<TClient, THook>>, EnvironmentMetadata, Logger)