--- title: "Bulk Create Digital Humans" method: POST path: "/v1/create-digital-humans" tags: ["Digital Humans"] --- # Bulk Create Digital Humans `POST /v1/create-digital-humans` Bulk create digital humans. All database operations are batched to O(1) round-trips regardless of input size. Items that fail validation are reported in `errors`; the rest proceed to creation. ## Headers - `X-API-Key` string, required ## Request body - BulkCreateDigitalHumansRequest — Request model for bulk creating digital humans. - `simulation_ids` integer[], nullable — list of simulation IDs to associate with all digital humans in this request - `digital_humans` DigitalHumanRequestData[], required — list of digital humans to create (min 1, max 100) - `bluejay_as_code_id` string, uuid, nullable — Stable code-addressable identifier - `intent` string — Description of the digital human. Leave empty when enriched_playback is set; the agent uses playback instead. - `success_criteria` string — Success criteria. Leave empty when enriched_playback is set; the agent uses playback instead. - `name` string, nullable — Name of the digital human - `tag` string, nullable — Tag for categorizing the digital human - `tags` string[], nullable — Full tag list (takes precedence over `tag`). For transcript replay, tags[0]='transcript_generated' and tags[1] in {'smart','strict'}. - `language` 'en' | 'es' | 'pt' | 'ja' | 'tr' | 'hi' | 'ar' | 'he' | 'ru' | 'zh' | 'ml' | 'fr' | 'yue' | 'vi' | 'de' | 'ko' | 'ur' | 'te' | 'ta' | 'mal' | 'kn' | 'mr' | 'gu' | 'ad' — Enum for supported language types. - `accent` 'multilingual' | 'american' | 'american2' | 'mature' | 'southern' | 'italian' | 'indian' | 'british' | 'australian' | 'scottish' | 'irish' | 'welsh' | 'mexican' | 'spanish' | 'portuguese' | 'french' | 'turkish' | 'japanese' | 'hindi' | 'arabic' | 'egyptian' | 'levantine' | 'hebrew' | 'russian' | 'chinese' | 'german' | 'korean' | 'urdu' | 'telugu' | 'tamil' | 'malayalam' | 'kannada' | 'marathi' | 'gujarati' | 'autodetect' — Enum for accent options. - `gender` 'male' | 'female' — Enum for gender options. - `custom_voice_id` string, nullable — ElevenLabs voice_id of a cloned voice. When set, the agent speaks with this voice instead of the accent's stock voice (language/accent still drive prompts and STT). - `background_noise` 'none' | 'office' | 'talking' | 'traffic' | 'cafe' | 'park' | 'tv' | 'noisy_restaurant' | 'hospital' | 'custom' — Enum for background noise options. - `custom_background_noise_url` string, nullable — Bucket-relative path of the uploaded custom background noise audio; required when background_noise is 'custom' - `voice_speed` 'slowest' | 'slow' | 'normal' | 'fast' | 'fastest' — Enum for voice speed options. - `audio_quality` 'high' | 'medium' | 'low' | 'horrible' — Enum for audio quality options. - `fluency` 'beginner' | 'intermediate' | 'native' — Enum for fluency level options. - `verbosity` 'low' | 'medium' | 'high' — Enum for verbosity level options. - `phone_number` string, nullable — Phone number for the digital human - `extension` string, nullable — Extension dialed as DTMF after connecting when this DH calls an inbound agent - `outbound_text_number` string, nullable — Outbound text number - `follow_up_sms_success_criteria` string, nullable — Criteria the captured follow-up SMS is graded against - `background_noise_volume` number, nullable — Volume of background noise - `expected_tool_calls` ExpectedToolCall[], nullable — Expected tool call outputs - `name` string, required — Name of the expected tool call - `parameters` object, nullable — Expected parameters passed to the tool call - `output` unknown - `allow_end_call_tool` boolean, nullable — Allow the digital human to end the tool call - `allow_silence_tool` boolean, nullable — Allow the digital human to use the silence tool - `allow_dtmf_tool` boolean, nullable — Allow the digital human to use the DTMF tool - `default_dtmf_or_voiced` 'voiced' | 'dtmf' — Enum for DTMF vs voiced input mode. - `silence_tool_instructions` string, nullable — Tool instructions; use "default" for built-in behavior or custom text - `endpointing_delay` number, nullable — Delay for endpointing - `creativity` number, nullable — How creative the digital human is (Model temperature) - `hangup_phrases` string[], nullable — Phrases that trigger hangup - `hangup_instructions` string, nullable — Freeform instructions for how/when to hang up - `silence_timeout` integer, nullable — Silence timeout in seconds - `role_description` string, nullable — Description of the role for the digital human - `traits` SrcModelsTraitsTraitTrait[], nullable — List of traits associated with this digital human - `trait_name` string, required — Name of the trait - `trait_data_type` 'BOOLEAN' | 'STRING' | 'DATE' | 'NUMBER', required — Enum for trait data types. - `value` unknown, required - `is_sip_header` boolean — Whether this trait is a SIP header - `interruptions` union — Interruption configuration for the digital human - SimpleInterruptionConfig — Simple interruption configuration with predefined levels. - `type` 'none' | 'low' | 'medium' | 'high', required — Interruption level: none, low, medium, or high - KeywordsInterruptionConfig — Keywords-based interruption configuration. - `type` 'keywords', required — Type discriminator for keywords interruption (required) - `keyword_responses` KeywordResponse[], required — Array of keyword response objects, each with keyword, response, and optional probability - `keyword` string, required — The keyword to listen for and interrupt on - `response` string, required — Response to say when keyword is detected - `probability` number — Probability of responding (0.0-1.0) - defaults to 1.0 - ContextualInterruptionConfig — Contextual/LLM-based interruption configuration. - `type` 'contextual', required — Type discriminator for contextual interruption - `contextual_prompt` string, required — Contextual prompt for LLM-based interruptions - `scripted_responses` ScriptedResponse[], nullable — List of scripted responses for the digital human - `match_type` 'exact' | 'context', required — Enum for scripted response match types. - `match_phrase` string, required — The phrase to match against - `response_type` 'phrase' | 'silence' | 'dtmf' | 'audio', required — Enum for scripted response types. - `response_value` string, nullable — The response text (required for 'phrase' and 'dtmf', optional for 'silence', unused for 'audio') - `scripted_response_audio_snippet_url` string, nullable — Skywatch bucket path of an uploaded audio file. AUDIO rows only — the agent downloads and plays this verbatim. - `occurrence_mode` 'always' | 'first_n', required — Enum for scripted response occurrence modes. - `occurrence_n` integer, nullable — Number of times to apply (required if occurrence_mode is 'first_n', must be > 0) - `silence_duration` integer, nullable — Duration in seconds for silence period (only used when response_type is 'silence', must be > 0 if provided) - `speaks_first_config` SpeaksFirstConfig — Configuration for whether the digital human speaks first and how. Stored as JSONB in Supabase (speaks_first_config). - `speaks_first` boolean, required — Whether the digital human speaks first in the call - `mode` 'custom' | 'ai_generated', nullable — When speaks_first is true: 'custom' (user-provided message) or 'ai_generated' (no message) - `message` string, nullable — When mode is 'custom', the message the digital human says first - `original_transcript` string, nullable — Original transcript text to replay. When provided (and formatted_transcript not provided), utterances are extracted via LLM and used as the intent. - `formatted_transcript` object[], nullable — Pre-computed structured transcript as [{role, utterance}]. When provided, skips the format-transcript LLM call. - `enriched_playback` object[], nullable — Workflow playbook v2 (JSON list). Speech steps use type "message" only. User: intent and/or exact_message. Agent: criteria and/or exact_match_text. - `workflow_v2_id` string, uuid, nullable — Source workflow_v2 UUID — set when this DH is generated from a workflow - `workflow_path_index` integer, nullable — 0-based path index within enumerate_workflow_v2_paths output - `tested_path` string[], nullable — Ordered names of the workflow path steps this DH was generated from (Generate-From-Workflow only). - `journey_steps` JourneyStep[], nullable - `step` integer, required - `intent` string - `success_criteria` string - `modality` 'voice' | 'sms' - `direction` 'inbound' | 'outbound', nullable - `wait_seconds` integer, nullable - `response_window_seconds` integer, nullable - `unresponsive` boolean - `always_on_mode` boolean, nullable — Whether always-on mode is enabled - `always_on_active` boolean, nullable — When true, this DH actively receives calls on phone_number. Defaults to True when phone_number is set. - `num_runs` integer, nullable — Number of times this digital human is run per simulation run (run count). - `livekit_metadata` object, nullable — LiveKit-specific configuration and metadata for this digital human - `test_name` string, nullable — User-facing label for this digital human - `attack_vector` string, nullable — Attack tactic key from VECTOR_CATALOG. NULL for non-red-team. - `attack_type` string, nullable — Umbrella attack category (recon | social_engineering | prompt_injection | audio_native). - `attack_plan` object, nullable — Crescendo escalation ladder {objective, breach_predicate, ladder}. ## Response `200` Successful Response - BulkCreateDigitalHumansResponse — Response model for bulk digital human creation. - `created` SrcModelsDigitalHumansDigitalHumanResponseDigitalHumanResponse[] — Successfully created digital humans - `digital_human` DigitalHumanResponseData, required — Response model for digital human data. - `id` integer, nullable — Unique identifier for the digital human - `bluejay_as_code_id` string, uuid, nullable — Stable code-addressable identifier - `intent` string, required — Description of the digital human - `success_criteria` string, required — Success criteria for the digital human - `tag` string, nullable — Tag for categorizing the digital human - `tags` string[], nullable — All tags associated with the digital human - `name` string, nullable — Name of the digital human - `language` 'en' | 'es' | 'pt' | 'ja' | 'tr' | 'hi' | 'ar' | 'he' | 'ru' | 'zh' | 'ml' | 'fr' | 'yue' | 'vi' | 'de' | 'ko' | 'ur' | 'te' | 'ta' | 'mal' | 'kn' | 'mr' | 'gu' | 'ad' — Enum for supported language types. - `accent` 'multilingual' | 'american' | 'american2' | 'mature' | 'southern' | 'italian' | 'indian' | 'british' | 'australian' | 'scottish' | 'irish' | 'welsh' | 'mexican' | 'spanish' | 'portuguese' | 'french' | 'turkish' | 'japanese' | 'hindi' | 'arabic' | 'egyptian' | 'levantine' | 'hebrew' | 'russian' | 'chinese' | 'german' | 'korean' | 'urdu' | 'telugu' | 'tamil' | 'malayalam' | 'kannada' | 'marathi' | 'gujarati' | 'autodetect' — Enum for accent options. - `gender` 'male' | 'female' — Enum for gender options. - `custom_voice_id` string, nullable — ElevenLabs voice_id of a cloned voice used for TTS, if assigned - `background_noise` string, nullable — Type of background noise - `custom_background_noise_url` string, nullable — Bucket-relative path of the uploaded custom background noise audio when background_noise is 'custom' - `voice_speed` string, nullable — Speed of the digital human's voice - `audio_quality` string, nullable — Audio quality of the digital human's voice - `fluency` string, nullable — Fluency level of the digital human's speech - `verbosity` string, nullable — Verbosity level of the digital human's responses - `phone_number` string, nullable — Phone number for the digital human - `extension` string, nullable — Extension dialed as DTMF after connecting when this DH calls an inbound agent - `outbound_text_number` string, nullable — Outbound text number - `follow_up_sms_success_criteria` string, nullable — Criteria the captured follow-up SMS is graded against - `websocket_url` string, nullable — WebSocket URL for real-time communication - `background_noise_volume` number, nullable — Volume of background noise - `expected_tool_calls` ExpectedToolCall[], nullable — Expected tool calls - `name` string, required — Name of the expected tool call - `parameters` object, nullable — Expected parameters passed to the tool call - `output` unknown - `allow_end_call_tool` boolean, nullable — Allow the digital human to end the tool call - `allow_silence_tool` boolean, nullable — Allow the digital human to use the silence tool - `allow_dtmf_tool` boolean, nullable — Allow the digital human to use the DTMF tool - `default_dtmf_or_voiced` string, nullable - `silence_tool_instructions` string, nullable — Tool instructions; "default" or custom text - `endpointing_delay` number, nullable — Delay for endpointing - `creativity` number, nullable — Temperature setting - `hangup_phrases` string[], nullable — Phrases that trigger hangup - `hangup_instructions` string, nullable — Freeform instructions for how/when to hang up - `silence_timeout` integer, nullable — Silence timeout in seconds - `role_description` string, nullable — Description of the role for the digital human - `created_at` string, date-time, nullable — When this digital human was created - `traits` TraitOutput[], nullable — List of traits associated with this digital human - `trait_name` string, required — Name of the trait - `trait_data_type` 'BOOLEAN' | 'STRING' | 'DATE' | 'NUMBER', required — Enum for trait data types. - `value` unknown, required - `is_sip_header` boolean — Whether this trait is a SIP header - `interruptions` object, nullable — Interruption configuration for the digital human - `scripted_responses` ScriptedResponse[], nullable — List of scripted responses for the digital human - `match_type` 'exact' | 'context', required — Enum for scripted response match types. - `match_phrase` string, required — The phrase to match against - `response_type` 'phrase' | 'silence' | 'dtmf' | 'audio', required — Enum for scripted response types. - `response_value` string, nullable — The response text (required for 'phrase' and 'dtmf', optional for 'silence', unused for 'audio') - `scripted_response_audio_snippet_url` string, nullable — Skywatch bucket path of an uploaded audio file. AUDIO rows only — the agent downloads and plays this verbatim. - `occurrence_mode` 'always' | 'first_n', required — Enum for scripted response occurrence modes. - `occurrence_n` integer, nullable — Number of times to apply (required if occurrence_mode is 'first_n', must be > 0) - `silence_duration` integer, nullable — Duration in seconds for silence period (only used when response_type is 'silence', must be > 0 if provided) - `speaks_first_config` SpeaksFirstConfig — Configuration for whether the digital human speaks first and how. Stored as JSONB in Supabase (speaks_first_config). - `speaks_first` boolean, required — Whether the digital human speaks first in the call - `mode` 'custom' | 'ai_generated', nullable — When speaks_first is true: 'custom' (user-provided message) or 'ai_generated' (no message) - `message` string, nullable — When mode is 'custom', the message the digital human says first - `original_transcript` string, nullable — The original transcript text attached to this digital human - `formatted_transcript` object[], nullable — Structured transcript as [{"role": "USER"|"AGENT", "utterance": "..."}] - `enriched_playback` object[], nullable — Optional enriched playback stored as JSONB: a list of turn objects - `workflow_v2_id` string, uuid, nullable — Source workflow_v2 UUID — set when this DH is generated from a workflow - `workflow_path_index` integer, nullable — 0-based path index within enumerate_workflow_v2_paths output - `tested_path` string[], nullable — Ordered names of the workflow path steps this DH was generated from (Generate-From-Workflow only). - `livekit_metadata` object, nullable — LiveKit-specific configuration and metadata for this digital human - `test_name` string, nullable — User-facing label for this digital human - `journey_steps` JourneyStep[], nullable — Ordered journey steps for a multi-step (Customer Journey) digital human - `step` integer, required - `intent` string - `success_criteria` string - `modality` 'voice' | 'sms' - `direction` 'inbound' | 'outbound', nullable - `wait_seconds` integer, nullable - `response_window_seconds` integer, nullable - `unresponsive` boolean - `simulation_ids` integer[], nullable — List of simulation IDs associated with this digital human - `simulation_id` integer, nullable — ID of the associated simulation. Use simulation_ids instead. - `errors` object[] — Errors for items that failed to create, with index and detail ## Other responses - `422` — Validation Error --- [API](https://skmtc.net/getbluejay/apis/bluejay-api.md) · [All operations](https://skmtc.net/getbluejay/apis/bluejay-api/llms.txt) · [OpenAPI document](https://skmtc-service-staging.skmtc.workers.dev/v1/apis/getbluejay/bluejay-api/versions/f48cef80963f/schema)