Phase 3

Author: rhurey

sdk/ai/Azure.AI.VoiceLive/src/ConversationSessionOptions.cs

@@ -0,0 +1,121 @@
+// Copyright (c) Microsoft Corporation. All rights reserved.
+// Licensed under the MIT License.
+
+using System;
+using System.Collections.Generic;
+using System.ComponentModel;
+using Azure.Core;
+
+namespace Azure.AI.VoiceLive
+{
+    /// <summary>
+    /// Represents configuration options for a VoiceLive conversation session.
+    /// </summary>
+    /// <remarks>
+    /// This class provides configuration options specifically tailored for conversational interactions
+    /// with the VoiceLive service, including voice selection, tool usage, and conversation management.
+    /// </remarks>
+    public class ConversationSessionOptions : VoiceLiveSessionOptions
+    {
+        /// <summary>
+        /// Gets or sets the voice configuration for the conversation.
+        /// </summary>
+        /// <value>
+        /// The voice configuration to use for generating spoken responses. If not specified, 
+        /// the service will use a default voice.
+        /// </value>
+        public BinaryData Voice { get; set; }
+
+        /// <summary>
+        /// Gets or sets the model to use for the conversation.
+        /// </summary>
+        /// <value>
+        /// The model identifier for conversation processing. If not specified, the service will use a default model.
+        /// </value>
+        public string Model { get; set; }
+
+        /// <summary>
+        /// Gets or sets the instructions for the conversation assistant.
+        /// </summary>
+        /// <value>
+        /// Instructions that guide the assistant's behavior and responses during the conversation.
+        /// </value>
+        public string Instructions { get; set; }
+
+        /// <summary>
+        /// Gets or sets the tools available to the conversation assistant.
+        /// </summary>
+        /// <value>
+        /// A list of tools that the assistant can use during the conversation.
+        /// </value>
+        public IList<VoiceLiveTool> Tools { get; set; } = new List<VoiceLiveTool>();
+
+        /// <summary>
+        /// Gets or sets the tool choice strategy for the conversation.
+        /// </summary>
+        /// <value>
+        /// Specifies how the assistant should choose which tools to use. If not specified,
+        /// the assistant will automatically decide when to use tools.
+        /// </value>
+        public BinaryData ToolChoice { get; set; }
+
+        /// <summary>
+        /// Gets or sets a value indicating whether to enable parallel tool calling.
+        /// </summary>
+        /// <value>
+        /// <c>true</c> to allow the assistant to call multiple tools in parallel; otherwise, <c>false</c>.
+        /// Default is <c>false</c>.
+        /// </value>
+        public bool ParallelToolCalls { get; set; }
+
+        /// <summary>
+        /// Initializes a new instance of the <see cref="ConversationSessionOptions"/> class.
+        /// </summary>
+        public ConversationSessionOptions() : base()
+        {
+        }
+
+        /// <summary>
+        /// Converts the conversation session options to a <see cref="VoiceLiveRequestSession"/> instance.
+        /// </summary>
+        /// <returns>A <see cref="VoiceLiveRequestSession"/> instance configured with the current options.</returns>
+        internal override VoiceLiveRequestSession ToRequestSession()
+        {
+            var session = base.ToRequestSession();
+
+            if (Voice != null)
+            {
+                session.Voice = Voice;
+            }
+
+            if (!string.IsNullOrEmpty(Model))
+            {
+                session.Model = Model;
+            }
+
+            if (!string.IsNullOrEmpty(Instructions))
+            {
+                session.Instructions = Instructions;
+            }
+
+            if (Tools != null && Tools.Count > 0)
+            {
+                session.Tools.Clear();
+                foreach (var tool in Tools)
+                {
+                    session.Tools.Add(tool);
+                }
+            }
+
+            if (ToolChoice != null)
+            {
+                session.ToolChoice = ToolChoice;
+            }
+
+            // Note: ParallelToolCalls is not supported in VoiceLiveRequestSession
+            // This would need to be handled differently or added to generated model
+
+            return session;
+        }
+    }
+}

sdk/ai/Azure.AI.VoiceLive/src/TranscriptionSessionOptions.cs

@@ -0,0 +1,130 @@
+// Copyright (c) Microsoft Corporation. All rights reserved.
+// Licensed under the MIT License.
+
+using System;
+using System.Collections.Generic;
+using System.ComponentModel;
+using Azure.Core;
+
+namespace Azure.AI.VoiceLive
+{
+    /// <summary>
+    /// Represents configuration options for a VoiceLive transcription session.
+    /// </summary>
+    /// <remarks>
+    /// This class provides configuration options specifically tailored for audio transcription
+    /// scenarios with the VoiceLive service, focusing on audio processing and transcription accuracy.
+    /// </remarks>
+    public class TranscriptionSessionOptions : VoiceLiveSessionOptions
+    {
+        /// <summary>
+        /// Gets or sets the language for transcription.
+        /// </summary>
+        /// <value>
+        /// The language code (e.g., "en-US", "fr-FR") to use for transcription.
+        /// If not specified, the service will attempt to auto-detect the language.
+        /// </value>
+        public string Language { get; set; }
+
+        /// <summary>
+        /// Gets or sets the transcription model to use.
+        /// </summary>
+        /// <value>
+        /// The model identifier for transcription processing. If not specified, 
+        /// the service will use a default transcription model.
+        /// </value>
+        public string Model { get; set; }
+
+        /// <summary>
+        /// Gets or sets a value indicating whether to include confidence scores in transcription results.
+        /// </summary>
+        /// <value>
+        /// <c>true</c> to include confidence scores for transcribed text; otherwise, <c>false</c>.
+        /// Default is <c>false</c>.
+        /// </value>
+        public bool IncludeConfidenceScores { get; set; }
+
+        /// <summary>
+        /// Gets or sets a value indicating whether to include timestamps in transcription results.
+        /// </summary>
+        /// <value>
+        /// <c>true</c> to include word-level timestamps in transcribed text; otherwise, <c>false</c>.
+        /// Default is <c>false</c>.
+        /// </value>
+        public bool IncludeTimestamps { get; set; }
+
+        /// <summary>
+        /// Gets or sets the audio noise reduction settings.
+        /// </summary>
+        /// <value>
+        /// Configuration for reducing noise in the input audio to improve transcription accuracy.
+        /// If not specified, default noise reduction settings will be used.
+        /// </value>
+        public VoiceLiveAudioNoiseReduction NoiseReduction { get; set; }
+
+        /// <summary>
+        /// Gets or sets the echo cancellation settings.
+        /// </summary>
+        /// <value>
+        /// Configuration for cancelling echo in the input audio to improve transcription accuracy.
+        /// If not specified, default echo cancellation settings will be used.
+        /// </value>
+        public VoiceLiveAudioEchoCancellation EchoCancellation { get; set; }
+
+        /// <summary>
+        /// Gets or sets a list of words or phrases to boost recognition accuracy.
+        /// </summary>
+        /// <value>
+        /// A list of domain-specific words or phrases that should be recognized more accurately.
+        /// This can improve transcription quality for specialized vocabulary.
+        /// </value>
+        public IList<string> CustomVocabulary { get; set; } = new List<string>();
+
+        /// <summary>
+        /// Initializes a new instance of the <see cref="TranscriptionSessionOptions"/> class.
+        /// </summary>
+        public TranscriptionSessionOptions() : base()
+        {
+            // Default modalities for transcription sessions - typically just audio
+            Modalities = new List<VoiceLiveModality> { VoiceLiveModality.Audio };
+        }
+
+        /// <summary>
+        /// Converts the transcription session options to a <see cref="VoiceLiveRequestSession"/> instance.
+        /// </summary>
+        /// <returns>A <see cref="VoiceLiveRequestSession"/> instance configured with the current options.</returns>
+        internal override VoiceLiveRequestSession ToRequestSession()
+        {
+            var session = base.ToRequestSession();
+
+            if (!string.IsNullOrEmpty(Language))
+            {
+                // Store language in additional properties since it might not be a direct property
+                session.AdditionalProperties["language"] = BinaryData.FromString($"\"{Language}\"");
+            }
+
+            if (!string.IsNullOrEmpty(Model))
+            {
+                session.Model = Model;
+            }
+
+            // Note: The following properties may need to be stored differently
+            // as they might not be direct properties on VoiceLiveRequestSession
+
+            if (NoiseReduction != null)
+            {
+                session.InputAudioNoiseReduction = NoiseReduction;
+            }
+
+            if (EchoCancellation != null)
+            {
+                session.InputAudioEchoCancellation = EchoCancellation;
+            }
+
+            // Custom vocabulary and other properties may need special handling
+            // as they may not be directly supported by VoiceLiveRequestSession
+
+            return session;
+        }
+    }
+}

sdk/ai/Azure.AI.VoiceLive/src/VoiceLiveResponseOptions.cs

@@ -0,0 +1,127 @@
+// Copyright (c) Microsoft Corporation. All rights reserved.
+// Licensed under the MIT License.
+
+using System;
+using System.Collections.Generic;
+using System.ComponentModel;
+using Azure.Core;
+
+namespace Azure.AI.VoiceLive
+{
+    /// <summary>
+    /// Represents configuration options for VoiceLive response generation.
+    /// </summary>
+    /// <remarks>
+    /// This class provides configuration options for controlling how the VoiceLive service
+    /// generates responses, including modalities, tools, and response formatting.
+    /// </remarks>
+    public class VoiceLiveResponseOptions
+    {
+        /// <summary>
+        /// Gets or sets the modalities to include in the response.
+        /// </summary>
+        /// <value>
+        /// A list of modalities (e.g., text, audio) that should be included in the response.
+        /// If not specified, the service will use default modalities.
+        /// </value>
+        public IList<VoiceLiveModality> Modalities { get; set; } = new List<VoiceLiveModality>();
+
+        /// <summary>
+        /// Gets or sets the instructions for response generation.
+        /// </summary>
+        /// <value>
+        /// Instructions that guide how the response should be generated.
+        /// </value>
+        public string Instructions { get; set; }
+
+        /// <summary>
+        /// Gets or sets the voice configuration for spoken responses.
+        /// </summary>
+        /// <value>
+        /// The voice configuration to use for generating spoken responses.
+        /// </value>
+        public BinaryData Voice { get; set; }
+
+        /// <summary>
+        /// Gets or sets the output audio format for the response.
+        /// </summary>
+        /// <value>
+        /// The audio format to use for output audio in the response.
+        /// </value>
+        public VoiceLiveAudioFormat? OutputAudioFormat { get; set; }
+
+        /// <summary>
+        /// Gets or sets the tools available during response generation.
+        /// </summary>
+        /// <value>
+        /// A list of tools that can be used during response generation.
+        /// </value>
+        public IList<VoiceLiveTool> Tools { get; set; } = new List<VoiceLiveTool>();
+
+        /// <summary>
+        /// Gets or sets the tool choice strategy for response generation.
+        /// </summary>
+        /// <value>
+        /// Specifies how tools should be chosen during response generation.
+        /// </value>
+        public string ToolChoice { get; set; }
+
+        /// <summary>
+        /// Gets or sets the temperature parameter for response generation.
+        /// </summary>
+        /// <value>
+        /// A value between 0.0 and 1.0 controlling the randomness of the response.
+        /// Higher values produce more random responses.
+        /// </value>
+        public float? Temperature { get; set; }
+
+        /// <summary>
+        /// Gets or sets the maximum number of tokens to generate in the response.
+        /// </summary>
+        /// <value>
+        /// The maximum number of tokens to generate. If not specified, the service will use a default limit.
+        /// </value>
+        public int? MaxOutputTokens { get; set; }
+
+        /// <summary>
+        /// Gets or sets a value indicating whether to commit the response to the conversation.
+        /// </summary>
+        /// <value>
+        /// <c>true</c> to commit the response to the conversation; otherwise, <c>false</c>.
+        /// Default is <c>true</c>.
+        /// </value>
+        public bool? Commit { get; set; }
+
+        /// <summary>
+        /// Gets or sets a value indicating whether to cancel any ongoing generation before starting this one.
+        /// </summary>
+        /// <value>
+        /// <c>true</c> to cancel ongoing generation; otherwise, <c>false</c>.
+        /// Default is <c>true</c>.
+        /// </value>
+        public bool? CancelPrevious { get; set; }
+
+        /// <summary>
+        /// Initializes a new instance of the <see cref="VoiceLiveResponseOptions"/> class.
+        /// </summary>
+        public VoiceLiveResponseOptions()
+        {
+        }
+
+        /// <summary>
+        /// Converts the response options to a <see cref="VoiceLiveResponseCreateParams"/> instance.
+        /// </summary>
+        /// <returns>A <see cref="VoiceLiveResponseCreateParams"/> instance configured with the current options.</returns>
+        /// <remarks>
+        /// This method uses the model factory to create the response parameters since the
+        /// VoiceLiveResponseCreateParams constructor is internal.
+        /// </remarks>
+        internal virtual VoiceLiveResponseCreateParams ToCreateParams()
+        {
+            // Since VoiceLiveResponseCreateParams has an internal constructor,
+            // we need to find another way to create it. For now, we'll return null
+            // and handle this at the call site.
+            return null;
+        }
+    }
+}