Class: Aws::LexRuntimeV2::AsyncClient

Inherits:
Seahorse::Client::AsyncBase show all
Includes:
AsyncClientStubs
Defined in:
gems/aws-sdk-lexruntimev2/lib/aws-sdk-lexruntimev2/async_client.rb

Instance Attribute Summary

Attributes inherited from Seahorse::Client::AsyncBase

#connection

Attributes inherited from Seahorse::Client::Base

#config, #handlers

API Operations collapse

Instance Method Summary collapse

Methods included from AsyncClientStubs

#send_events

Methods included from ClientStubs

#api_requests, #stub_data, #stub_responses

Methods inherited from Seahorse::Client::AsyncBase

#close_connection, #connection_errors, #new_connection, #operation_names

Methods inherited from Seahorse::Client::Base

add_plugin, api, clear_plugins, define, new, #operation_names, plugins, remove_plugin, set_api, set_plugins

Methods included from Seahorse::Client::HandlerBuilder

#handle, #handle_request, #handle_response

Constructor Details

#initialize(*args) ⇒ AsyncClient

@option options [required, Aws::CredentialProvider] :credentials Your AWS credentials. This can be an instance of any one of the following classes:

* `Aws::Credentials` - Used for configuring static, non-refreshing
  credentials.

* `Aws::SharedCredentials` - Used for loading static credentials from a
  shared file, such as `~/.aws/config`.

* `Aws::AssumeRoleCredentials` - Used when you need to assume a role.

* `Aws::AssumeRoleWebIdentityCredentials` - Used when you need to
  assume a role after providing credentials via the web.

* `Aws::SSOCredentials` - Used for loading credentials from AWS SSO using an
  access token generated from `aws login`.

* `Aws::ProcessCredentials` - Used for loading credentials from a
  process that outputs to stdout.

* `Aws::InstanceProfileCredentials` - Used for loading credentials
  from an EC2 IMDS on an EC2 instance.

* `Aws::ECSCredentials` - Used for loading credentials from
  instances running in ECS.

* `Aws::CognitoIdentityCredentials` - Used for loading credentials
  from the Cognito Identity service.

When `:credentials` are not configured directly, the following
locations will be searched for credentials:

* `Aws.config[:credentials]`
* The `:access_key_id`, `:secret_access_key`, and `:session_token` options.
* ENV['AWS_ACCESS_KEY_ID'], ENV['AWS_SECRET_ACCESS_KEY']
* `~/.aws/credentials`
* `~/.aws/config`
* EC2/ECS IMDS instance profile - When used by default, the timeouts
  are very aggressive. Construct and pass an instance of
  `Aws::InstanceProfileCredentails` or `Aws::ECSCredentials` to
  enable retries and extended timeouts. Instance profile credential
  fetching can be disabled by setting ENV['AWS_EC2_METADATA_DISABLED']
  to true.

@option options [required, String] :region The AWS region to connect to. The configured :region is used to determine the service :endpoint. When not passed, a default :region is searched for in the following locations:

* `Aws.config[:region]`
* `ENV['AWS_REGION']`
* `ENV['AMAZON_REGION']`
* `ENV['AWS_DEFAULT_REGION']`
* `~/.aws/credentials`
* `~/.aws/config`

@option options [String] :access_key_id

@option options [Boolean] :adaptive_retry_wait_to_fill (true) Used only in adaptive retry mode. When true, the request will sleep until there is sufficent client side capacity to retry the request. When false, the request will raise a RetryCapacityNotAvailableError and will not retry instead of sleeping.

@option options [Boolean] :convert_params (true) When true, an attempt is made to coerce request parameters into the required types.

@option options [Boolean] :correct_clock_skew (true) Used only in standard and adaptive retry modes. Specifies whether to apply a clock skew correction and retry requests with skewed client clocks.

@option options [String] :defaults_mode ("legacy") See DefaultsModeConfiguration for a list of the accepted modes and the configuration defaults that are included.

@option options [Boolean] :disable_request_compression (false) When set to 'true' the request body will not be compressed for supported operations.

@option options [String] :endpoint The client endpoint is normally constructed from the :region option. You should only configure an :endpoint when connecting to test or custom endpoints. This should be a valid HTTP(S) URI.

@option options [Proc] :event_stream_handler When an EventStream or Proc object is provided, it will be used as callback for each chunk of event stream response received along the way.

@option options [Boolean] :ignore_configured_endpoint_urls Setting to true disables use of endpoint URLs provided via environment variables and the shared configuration file.

@option options [Proc] :input_event_stream_handler When an EventStream or Proc object is provided, it can be used for sending events for the event stream.

@option options [Aws::Log::Formatter] :log_formatter (Aws::Log::Formatter.default) The log formatter.

@option options [Symbol] :log_level (:info) The log level to send messages to the :logger at.

@option options [Logger] :logger The Logger instance to send log messages to. If this option is not set, logging will be disabled.

@option options [Integer] :max_attempts (3) An integer representing the maximum number attempts that will be made for a single request, including the initial attempt. For example, setting this value to 5 will result in a request being retried up to 4 times. Used in standard and adaptive retry modes.

@option options [Proc] :output_event_stream_handler When an EventStream or Proc object is provided, it will be used as callback for each chunk of event stream response received along the way.

@option options [String] :profile ("default") Used when loading credentials from the shared credentials file at HOME/.aws/credentials. When not specified, 'default' is used.

@option options [Integer] :request_min_compression_size_bytes (10240) The minimum size in bytes that triggers compression for request bodies. The value must be non-negative integer value between 0 and 10485780 bytes inclusive.

@option options [Proc] :retry_backoff A proc or lambda used for backoff. Defaults to 2**retries * retry_base_delay. This option is only used in the legacy retry mode.

@option options [Float] :retry_base_delay (0.3) The base delay in seconds used by the default backoff function. This option is only used in the legacy retry mode.

@option options [Symbol] :retry_jitter (:none) A delay randomiser function used by the default backoff function. Some predefined functions can be referenced by name - :none, :equal, :full, otherwise a Proc that takes and returns a number. This option is only used in the legacy retry mode.

@see https://www.awsarchitectureblog.com/2015/03/backoff.html

@option options [Integer] :retry_limit (3) The maximum number of times to retry failed requests. Only ~ 500 level server errors and certain ~ 400 level client errors are retried. Generally, these are throttling errors, data checksum errors, networking errors, timeout errors, auth errors, endpoint discovery, and errors from expired credentials. This option is only used in the legacy retry mode.

@option options [Integer] :retry_max_delay (0) The maximum number of seconds to delay between retries (0 for no limit) used by the default backoff function. This option is only used in the legacy retry mode.

@option options [String] :retry_mode ("legacy") Specifies which retry algorithm to use. Values are:

* `legacy` - The pre-existing retry behavior.  This is default value if
  no retry mode is provided.

* `standard` - A standardized set of retry rules across the AWS SDKs.
  This includes support for retry quotas, which limit the number of
  unsuccessful retries a client can make.

* `adaptive` - An experimental retry mode that includes all the
  functionality of `standard` mode along with automatic client side
  throttling.  This is a provisional mode that may change behavior
  in the future.

@option options [String] :sdk_ua_app_id A unique and opaque application ID that is appended to the User-Agent header as app/. It should have a maximum length of 50.

@option options [String] :secret_access_key

@option options [String] :session_token

@option options [Boolean] :stub_responses (false) Causes the client to return stubbed responses. By default fake responses are generated and returned. You can specify the response data to return or errors to raise by calling ClientStubs#stub_responses. See ClientStubs for more information.

** Please note ** When response stubbing is enabled, no HTTP
requests are made, and retries are disabled.

@option options [Aws::TokenProvider] :token_provider A Bearer Token Provider. This can be an instance of any one of the following classes:

* `Aws::StaticTokenProvider` - Used for configuring static, non-refreshing
  tokens.

* `Aws::SSOTokenProvider` - Used for loading tokens from AWS SSO using an
  access token generated from `aws login`.

When `:token_provider` is not configured directly, the `Aws::TokenProviderChain`
will be used to search for tokens configured for your profile in shared configuration files.

@option options [Boolean] :use_dualstack_endpoint When set to true, dualstack enabled endpoints (with .aws TLD) will be used if available.

@option options [Boolean] :use_fips_endpoint When set to true, fips compatible endpoints will be used if available. When a fips region is used, the region is normalized and this config is set to true.

@option options [Boolean] :validate_params (true) When true, request parameters are validated before sending the request.

@option options [Aws::LexRuntimeV2::EndpointProvider] :endpoint_provider The endpoint provider used to resolve endpoints. Any object that responds to #resolve_endpoint(parameters) where parameters is a Struct similar to Aws::LexRuntimeV2::EndpointParameters



285
286
287
288
289
290
# File 'gems/aws-sdk-lexruntimev2/lib/aws-sdk-lexruntimev2/async_client.rb', line 285

def initialize(*args)
  unless Kernel.const_defined?("HTTP2")
    raise "Must include http/2 gem to use AsyncClient instances."
  end
  super
end

Instance Method Details

#start_conversation(params = {}) ⇒ Types::StartConversationResponse

Starts an HTTP/2 bidirectional event stream that enables you to send audio, text, or DTMF input in real time. After your application starts a conversation, users send input to Amazon Lex V2 as a stream of events. Amazon Lex V2 processes the incoming events and responds with streaming text or audio events.

Audio input must be in the following format: audio/lpcm sample-rate=8000 sample-size-bits=16 channel-count=1; is-big-endian=false.

If the optional post-fulfillment response is specified, the messages are returned as follows. For more information, see PostFulfillmentStatusSpecification.

  • Success message - Returned if the Lambda function completes successfully and the intent state is fulfilled or ready fulfillment if the message is present.

  • Failed message - The failed message is returned if the Lambda function throws an exception or if the Lambda function returns a failed intent state without a message.

  • Timeout message - If you don't configure a timeout message and a timeout, and the Lambda function doesn't return within 30 seconds, the timeout message is returned. If you configure a timeout, the timeout message is returned when the period times out.

For more information, see Completion message.

If the optional update message is configured, it is played at the specified frequency while the Lambda function is running and the update message state is active. If the fulfillment update message is not active, the Lambda function runs with a 30 second timeout.

For more information, see Update message

The StartConversation operation is supported only in the following SDKs:

Examples:

Bi-directional EventStream Operation Example


You can signal input events after initial request is
established, events will be sent to stream
immediately (once stream connection is established successfully).

To signal events, you can call #signal methods from an Aws::LexRuntimeV2::EventStreams::StartConversationRequestEventStream object.
Make sure signal events before calling #wait or #join! at async response.

  input_stream = Aws::LexRuntimeV2::EventStreams::StartConversationRequestEventStream.new

  async_resp = client.start_conversation( # params input,
    input_event_stream_handler: input_stream) do |out_stream|

    # register callbacks for events arrival
    out_stream.on_playback_interruption_event_event do |event|
      event # => Aws::LexRuntimeV2::Types::PlaybackInterruptionEvent
    end
    out_stream.on_transcript_event_event do |event|
      event # => Aws::LexRuntimeV2::Types::TranscriptEvent
    end
    out_stream.on_intent_result_event_event do |event|
      event # => Aws::LexRuntimeV2::Types::IntentResultEvent
    end
    out_stream.on_text_response_event_event do |event|
      event # => Aws::LexRuntimeV2::Types::TextResponseEvent
    end
    out_stream.on_audio_response_event_event do |event|
      event # => Aws::LexRuntimeV2::Types::AudioResponseEvent
    end
    out_stream.on_heartbeat_event_event do |event|
      event # => Aws::LexRuntimeV2::Types::HeartbeatEvent
    end
    out_stream.on_access_denied_exception_event do |event|
      event # => Aws::LexRuntimeV2::Types::AccessDeniedException
    end
    out_stream.on_resource_not_found_exception_event do |event|
      event # => Aws::LexRuntimeV2::Types::ResourceNotFoundException
    end
    out_stream.on_validation_exception_event do |event|
      event # => Aws::LexRuntimeV2::Types::ValidationException
    end
    out_stream.on_throttling_exception_event do |event|
      event # => Aws::LexRuntimeV2::Types::ThrottlingException
    end
    out_stream.on_internal_server_exception_event do |event|
      event # => Aws::LexRuntimeV2::Types::InternalServerException
    end
    out_stream.on_conflict_exception_event do |event|
      event # => Aws::LexRuntimeV2::Types::ConflictException
    end
    out_stream.on_dependency_failed_exception_event do |event|
      event # => Aws::LexRuntimeV2::Types::DependencyFailedException
    end
    out_stream.on_bad_gateway_exception_event do |event|
      event # => Aws::LexRuntimeV2::Types::BadGatewayException
    end

  end
  # => returns Aws::Seahorse::Client::AsyncResponse

  # signal events
  input_stream.signal_configuration_event_event( ... )
  input_stream.signal_audio_input_event_event( ... )
  input_stream.signal_dtmf_input_event_event( ... )
  input_stream.signal_text_input_event_event( ... )
  input_stream.signal_playback_completion_event_event( ... )
  input_stream.signal_disconnection_event_event( ... )

  # make sure signaling :end_stream in the end
  input_stream.signal_end_stream

  # wait until stream is closed before finalizing sync response
  resp = async_resp.wait
  # Or close stream and finalizing sync response immediately
  # resp = async_resp.join!

Inorder to streamingly processing events received, you can also provide an Aws::LexRuntimeV2::EventStreams::StartConversationResponseEventStream
object to register callbacks before initializing request instead of processing from request block

  output_stream = Aws::LexRuntimeV2::EventStreams::StartConversationResponseEventStream.new
  # register callbacks for events arrival
  output_stream.on_playback_interruption_event_event do |event|
    event # => Aws::LexRuntimeV2::Types::PlaybackInterruptionEvent
  end
  output_stream.on_transcript_event_event do |event|
    event # => Aws::LexRuntimeV2::Types::TranscriptEvent
  end
  output_stream.on_intent_result_event_event do |event|
    event # => Aws::LexRuntimeV2::Types::IntentResultEvent
  end
  output_stream.on_text_response_event_event do |event|
    event # => Aws::LexRuntimeV2::Types::TextResponseEvent
  end
  output_stream.on_audio_response_event_event do |event|
    event # => Aws::LexRuntimeV2::Types::AudioResponseEvent
  end
  output_stream.on_heartbeat_event_event do |event|
    event # => Aws::LexRuntimeV2::Types::HeartbeatEvent
  end
  output_stream.on_access_denied_exception_event do |event|
    event # => Aws::LexRuntimeV2::Types::AccessDeniedException
  end
  output_stream.on_resource_not_found_exception_event do |event|
    event # => Aws::LexRuntimeV2::Types::ResourceNotFoundException
  end
  output_stream.on_validation_exception_event do |event|
    event # => Aws::LexRuntimeV2::Types::ValidationException
  end
  output_stream.on_throttling_exception_event do |event|
    event # => Aws::LexRuntimeV2::Types::ThrottlingException
  end
  output_stream.on_internal_server_exception_event do |event|
    event # => Aws::LexRuntimeV2::Types::InternalServerException
  end
  output_stream.on_conflict_exception_event do |event|
    event # => Aws::LexRuntimeV2::Types::ConflictException
  end
  output_stream.on_dependency_failed_exception_event do |event|
    event # => Aws::LexRuntimeV2::Types::DependencyFailedException
  end
  output_stream.on_bad_gateway_exception_event do |event|
    event # => Aws::LexRuntimeV2::Types::BadGatewayException
  end
  output_stream.on_error_event do |event|
    # catch unmodeled error event in the stream
    raise event
    # => Aws::Errors::EventError
    # event.event_type => :error
    # event.error_code => String
    # event.error_message => String
  end

  async_resp = client.start_conversation ( #params input,
    input_event_stream_handler: input_stream
    output_event_stream_handler: output_stream
  )

  resp = async_resp.wait!

Besides above usage patterns for process events when they arrive immediately, you can also
iterate through events after response complete.

Events are available at resp.response_event_stream # => Enumerator

Request syntax with placeholder values


async_resp = async_client.start_conversation({
  bot_id: "BotIdentifier", # required
  bot_alias_id: "BotAliasIdentifier", # required
  locale_id: "LocaleId", # required
  session_id: "SessionId", # required
  conversation_mode: "AUDIO", # accepts AUDIO, TEXT
  input_event_stream_hander: EventStreams::StartConversationRequestEventStream.new,
})
# => Seahorse::Client::AsyncResponse
async_resp.wait
# => Seahorse::Client::Response
# Or use async_resp.join!

Response structure


All events are available at resp.response_event_stream:
resp.response_event_stream #=> Enumerator
resp.response_event_stream.event_types #=> [:playback_interruption_event, :transcript_event, :intent_result_event, :text_response_event, :audio_response_event, :heartbeat_event, :access_denied_exception, :resource_not_found_exception, :validation_exception, :throttling_exception, :internal_server_exception, :conflict_exception, :dependency_failed_exception, :bad_gateway_exception]

For :playback_interruption_event event available at #on_playback_interruption_event_event callback and response eventstream enumerator:
event.event_reason #=> String, one of "DTMF_START_DETECTED", "TEXT_DETECTED", "VOICE_START_DETECTED"
event.caused_by_event_id #=> String
event.event_id #=> String

For :transcript_event event available at #on_transcript_event_event callback and response eventstream enumerator:
event.transcript #=> String
event.event_id #=> String

For :intent_result_event event available at #on_intent_result_event_event callback and response eventstream enumerator:
event.input_mode #=> String, one of "Text", "Speech", "DTMF"
event.interpretations #=> Array
event.interpretations[0].nlu_confidence.score #=> Float
event.interpretations[0].sentiment_response.sentiment #=> String, one of "MIXED", "NEGATIVE", "NEUTRAL", "POSITIVE"
event.interpretations[0].sentiment_response.sentiment_score.positive #=> Float
event.interpretations[0].sentiment_response.sentiment_score.negative #=> Float
event.interpretations[0].sentiment_response.sentiment_score.neutral #=> Float
event.interpretations[0].sentiment_response.sentiment_score.mixed #=> Float
event.interpretations[0].intent.name #=> String
event.interpretations[0].intent.slots #=> Hash
event.interpretations[0].intent.slots["NonEmptyString"].value.original_value #=> String
event.interpretations[0].intent.slots["NonEmptyString"].value.interpreted_value #=> String
event.interpretations[0].intent.slots["NonEmptyString"].value.resolved_values #=> Array
event.interpretations[0].intent.slots["NonEmptyString"].value.resolved_values[0] #=> String
event.interpretations[0].intent.slots["NonEmptyString"].shape #=> String, one of "Scalar", "List", "Composite"
event.interpretations[0].intent.slots["NonEmptyString"].values #=> Array
event.interpretations[0].intent.slots["NonEmptyString"].values[0] #=> Types::Slot
event.interpretations[0].intent.slots["NonEmptyString"].sub_slots #=> Types::Slots
event.interpretations[0].intent.state #=> String, one of "Failed", "Fulfilled", "InProgress", "ReadyForFulfillment", "Waiting", "FulfillmentInProgress"
event.interpretations[0].intent.confirmation_state #=> String, one of "Confirmed", "Denied", "None"
event.interpretations[0].interpretation_source #=> String, one of "Bedrock", "Lex"
event.session_state.dialog_action.type #=> String, one of "Close", "ConfirmIntent", "Delegate", "ElicitIntent", "ElicitSlot", "None"
event.session_state.dialog_action.slot_to_elicit #=> String
event.session_state.dialog_action.slot_elicitation_style #=> String, one of "Default", "SpellByLetter", "SpellByWord"
event.session_state.dialog_action.sub_slot_to_elicit.name #=> String
event.session_state.dialog_action.sub_slot_to_elicit.sub_slot_to_elicit #=> Types::ElicitSubSlot
event.session_state.intent.name #=> String
event.session_state.intent.slots #=> Hash
event.session_state.intent.slots["NonEmptyString"].value.original_value #=> String
event.session_state.intent.slots["NonEmptyString"].value.interpreted_value #=> String
event.session_state.intent.slots["NonEmptyString"].value.resolved_values #=> Array
event.session_state.intent.slots["NonEmptyString"].value.resolved_values[0] #=> String
event.session_state.intent.slots["NonEmptyString"].shape #=> String, one of "Scalar", "List", "Composite"
event.session_state.intent.slots["NonEmptyString"].values #=> Array
event.session_state.intent.slots["NonEmptyString"].values[0] #=> Types::Slot
event.session_state.intent.slots["NonEmptyString"].sub_slots #=> Types::Slots
event.session_state.intent.state #=> String, one of "Failed", "Fulfilled", "InProgress", "ReadyForFulfillment", "Waiting", "FulfillmentInProgress"
event.session_state.intent.confirmation_state #=> String, one of "Confirmed", "Denied", "None"
event.session_state.active_contexts #=> Array
event.session_state.active_contexts[0].name #=> String
event.session_state.active_contexts[0].time_to_live.time_to_live_in_seconds #=> Integer
event.session_state.active_contexts[0].time_to_live.turns_to_live #=> Integer
event.session_state.active_contexts[0].context_attributes #=> Hash
event.session_state.active_contexts[0].context_attributes["ParameterName"] #=> String
event.session_state.session_attributes #=> Hash
event.session_state.session_attributes["NonEmptyString"] #=> String
event.session_state.originating_request_id #=> String
event.session_state.runtime_hints.slot_hints #=> Hash
event.session_state.runtime_hints.slot_hints["Name"] #=> Hash
event.session_state.runtime_hints.slot_hints["Name"]["Name"].runtime_hint_values #=> Array
event.session_state.runtime_hints.slot_hints["Name"]["Name"].runtime_hint_values[0].phrase #=> String
event.session_state.runtime_hints.slot_hints["Name"]["Name"].sub_slot_hints #=> Types::SlotHintsSlotMap
event.request_attributes #=> Hash
event.request_attributes["NonEmptyString"] #=> String
event.session_id #=> String
event.event_id #=> String
event.recognized_bot_member.bot_id #=> String
event.recognized_bot_member.bot_name #=> String

For :text_response_event event available at #on_text_response_event_event callback and response eventstream enumerator:
event.messages #=> Array
event.messages[0].content #=> String
event.messages[0].content_type #=> String, one of "CustomPayload", "ImageResponseCard", "PlainText", "SSML"
event.messages[0].image_response_card.title #=> String
event.messages[0].image_response_card.subtitle #=> String
event.messages[0].image_response_card.image_url #=> String
event.messages[0].image_response_card.buttons #=> Array
event.messages[0].image_response_card.buttons[0].text #=> String
event.messages[0].image_response_card.buttons[0].value #=> String
event.event_id #=> String

For :audio_response_event event available at #on_audio_response_event_event callback and response eventstream enumerator:
event.audio_chunk #=> String
event.content_type #=> String
event.event_id #=> String

For :heartbeat_event event available at #on_heartbeat_event_event callback and response eventstream enumerator:
event.event_id #=> String

For :access_denied_exception event available at #on_access_denied_exception_event callback and response eventstream enumerator:
event.message #=> String

For :resource_not_found_exception event available at #on_resource_not_found_exception_event callback and response eventstream enumerator:
event.message #=> String

For :validation_exception event available at #on_validation_exception_event callback and response eventstream enumerator:
event.message #=> String

For :throttling_exception event available at #on_throttling_exception_event callback and response eventstream enumerator:
event.message #=> String

For :internal_server_exception event available at #on_internal_server_exception_event callback and response eventstream enumerator:
event.message #=> String

For :conflict_exception event available at #on_conflict_exception_event callback and response eventstream enumerator:
event.message #=> String

For :dependency_failed_exception event available at #on_dependency_failed_exception_event callback and response eventstream enumerator:
event.message #=> String

For :bad_gateway_exception event available at #on_bad_gateway_exception_event callback and response eventstream enumerator:
event.message #=> String

Parameters:

  • params (Hash) (defaults to: {})

    ({})

Options Hash (params):

  • :bot_id (required, String)

    The identifier of the bot to process the request.

  • :bot_alias_id (required, String)

    The alias identifier in use for the bot that processes the request.

  • :locale_id (required, String)

    The locale where the session is in use.

  • :session_id (required, String)

    The identifier of the user session that is having the conversation.

  • :conversation_mode (String)

    The conversation type that you are using the Amazon Lex V2. If the conversation mode is AUDIO you can send both audio and DTMF information. If the mode is TEXT you can only send text.

Yields:

  • (output_event_stream_handler)

Returns:

See Also:



652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
# File 'gems/aws-sdk-lexruntimev2/lib/aws-sdk-lexruntimev2/async_client.rb', line 652

def start_conversation(params = {}, options = {}, &block)
  params = params.dup
  input_event_stream_handler = _event_stream_handler(
    :input,
    params.delete(:input_event_stream_handler),
    EventStreams::StartConversationRequestEventStream
  )
  output_event_stream_handler = _event_stream_handler(
    :output,
    params.delete(:output_event_stream_handler) || params.delete(:event_stream_handler),
    EventStreams::StartConversationResponseEventStream
  )

  yield(output_event_stream_handler) if block_given?

  req = build_request(:start_conversation, params)

  req.context[:input_event_stream_handler] = input_event_stream_handler
  req.handlers.add(Aws::Binary::EncodeHandler, priority: 55)
  req.context[:output_event_stream_handler] = output_event_stream_handler
  req.handlers.add(Aws::Binary::DecodeHandler, priority: 55)

  req.send_request(options, &block)
end