meeting-assistant/MeetingAssistant/Speakers/SpeakerIdentificationModels.cs

using MeetingAssistant.MeetingNotes;
using MeetingAssistant.Transcription;

namespace MeetingAssistant.Speakers;

public sealed record SpeakerIdentificationRequest(
    string AudioPath,
    MeetingNote MeetingNote,
    IReadOnlyList<TranscriptionSegment> Segments,
    IReadOnlyList<SpeakerAudioSample>? Samples = null,
    IReadOnlyDictionary<string, string>? KnownSpeakerMappings = null);

public sealed record SpeakerAudioSample(
    string Speaker,
    TranscriptionSegment Segment,
    byte[] WavBytes,
    double Score);

public sealed record SpeakerIdentificationResult(
    IReadOnlyList<TranscriptionSegment> Segments,
    IReadOnlyDictionary<string, string> SpeakerMappings,
    IReadOnlyList<SpeakerIdentityAttendeeMatch>? AttendeeMatches = null);

public sealed record SpeakerIdentityAttendeeMatch(
    string DisplayName,
    IReadOnlyList<string> AcceptedNames);

public sealed record SpeakerIdentityMatchRequest(
    string DiarizedSpeaker,
    byte[] UnknownSnippet,
    IReadOnlyList<SpeakerIdentityMatchCandidate> Candidates);

public sealed record SpeakerIdentityMatchCandidate(
    int IdentityId,
    string? CanonicalName,
    int ReferenceCount,
    IReadOnlyList<byte[]> Snippets);

public sealed record SpeakerIdentityMatch(int IdentityId);

public sealed record SpeakerIdentityMatchValidationRequest(
    string DiarizedSpeaker,
    int IdentityId,
    byte[] UnknownSnippet,
    IReadOnlyList<byte[]> KnownSnippets);

public interface ISpeakerIdentityMatcher
{
    Task<SpeakerIdentityMatch?> MatchAsync(
        SpeakerIdentityMatchRequest request,
        CancellationToken cancellationToken);
}

public interface ISpeakerIdentityMatchValidator
{
    Task<bool> ValidateSampleAsync(
        byte[] wavBytes,
        CancellationToken cancellationToken);

    Task<bool> ValidateMatchAsync(
        SpeakerIdentityMatchValidationRequest request,
        CancellationToken cancellationToken);
}

public interface ISpeakerSnippetExtractor
{
    Task<byte[]> ExtractSnippetAsync(
        string audioPath,
        IReadOnlyList<TranscriptionSegment> speakerSegments,
        CancellationToken cancellationToken);
}

public interface ISpeakerIdentificationService
{
    Task<SpeakerIdentificationResult> IdentifyKnownSpeakersAsync(
        SpeakerIdentificationRequest request,
        CancellationToken cancellationToken);

    Task<SpeakerIdentificationResult> IdentifyFinishedSpeakersAsync(
        SpeakerIdentificationRequest request,
        CancellationToken cancellationToken)
    {
        return IdentifyKnownSpeakersAsync(request, cancellationToken);
    }

    Task ApplySpeakerOverrideAsync(
        SpeakerIdentificationRequest request,
        string sourceSpeaker,
        string targetSpeaker,
        CancellationToken cancellationToken)
    {
        return Task.CompletedTask;
    }

    Task DeleteSpeakerIdentityAsync(
        string identity,
        CancellationToken cancellationToken)
    {
        return Task.CompletedTask;
    }

    Task<SpeakerIdentificationResult> ProcessFinishedTranscriptAsync(
        SpeakerIdentificationRequest request,
        CancellationToken cancellationToken);
}