Push metrics using Kafka Exporter with OpenTelemetry

I am trying to collect some metrics and export it to Kafka using OpenTelemetry. For a simple POC, I have a local file in the OpenMetrics format and I wish to parse it and push the metrics on Kafka.

I understand I should use kafkaexporter but I am not able to understand how the end-to-end flow would look like. A code snippet would be really helpful.

I am creating the Kafka exporter using the following code:

func newExporter(ctx context.Context) (exporter.Metrics, error) {
    f := kafkaexporter.NewFactory(kafkaexporter.WithMetricsMarshalers())
    cfg := f.CreateDefaultConfig()
    ts := component.TelemetrySettings{
        Logger:          logger,
        MeterProvider:   meterProvider(),
        MetricsLevel:    configtelemetry.LevelNormal,
    }
    cs := exporter.CreateSettings{
        ID:               component.NewID("metrics"),
        TelemetrySettings: ts,
        BuildInfo:        component.NewDefaultBuildInfo(),
    }

    return f.CreateMetricsExporter(ctx, cs, cfg)
}

func meterProvider() *metric.MeterProvider {
    return metric.NewMeterProvider()
}

Now say I have a few counters / gauge meters, how do I push it using the above exporter?

I was able to run the end-to-end flow where I collect some metrics from a downstream dependency, parse it and push to Kafka.

Here's the code snippet:

func collectMetrics() (pmetric.ResourceMetrics, error) {
    // collect and return metrics
}

func CollectMetrics(ctx context.Context) error {
    me, err := newExporter(ctx)
    if err != nil {
        return err
    }

    err = me.Start(ctx, nil)
    if err != nil {
        return err
    }
    defer me.Shutdown(ctx)

    ms := pmetric.NewMetrics()
    rm, err := collectMetrics()
    r := ms.ResourceMetrics().AppendEmpty()
    rm.CopyTo(r)

    err = me.ConsumeMetrics(ctx, ms)
    if err != nil {
        return err
    }

    return nil
}

The complete POC can be found here.