Prompted voice collection format for commands, corrections, confirmations, multi-turn instructions, and voice-agent instruction-following evaluation.
Instruction and Command Voice Sessions are built for teams evaluating how voice systems handle commands, follow-up instructions, corrections, confirmations, and clarification requests. The format is designed for intent classification, voice-agent testing, instruction-following evaluation, and command recognition.
Each session is structured around prompt categories such as task commands, device commands, agent-style requests, corrections, confirmations, and multi-turn instruction chains.
Structured collection
Instruction and Command Voice Sessions use structured prompt sets rather than open speech capture.
Metadata depth
The format supports utterance-level alignment for command recognition and optional word-level timing for ASR or…
Delivery-ready package
Includes audio clarity, prompt match, intent label reviews, plus dataset manifest, CSV catalog, consent status metadata, QA…
Request sample access
Share your command types, intent taxonomy, transcript depth, language requirements, and evaluation goals. We’ll reply with a technical scope for the collection.
Command audio files · Multi-turn instruction recordings · Correction markers · Confirmation markers · Clarification request markers
Prompt text · Prompt IDs · Transcript files · Intent category labels · Speaker metadata
Device metadata · Language metadata · Recording context metadata
Instruction and Command Voice Sessions use structured prompt sets rather than open speech capture. Contributors record commands, follow-up instructions, corrections, confirmations, and clarifications under defined task rules.
Audio clarity review · Prompt match review · Intent label review · Transcript alignment review · Correction marker review · Consent status review · Delivery approval
Audio files · Prompt JSON · Transcript JSON · Intent metadata CSV · Speaker/session metadata CSV · Dataset manifest JSON · CSV file catalog · QA summary · Consent status metadata · Delivery notes
Multi-speaker English dialogue sets with speaker-labeled transcripts, timestamped turns, accent/session metadata, optional separated-channel audio, and buyer-scoped annotations for ASR, diarization, and voice-agent evaluation.
Accent-focused speech sets with speaker profile metadata, region fields, prompt categories, transcript alignment, and QA status fields for ASR robustness and voice-agent coverage.
Styled speech sets with prompt IDs, transcript alignment, style labels, speaker metadata, audio quality fields, and buyer-scoped expression categories for TTS and voice model evaluation.
Role-based support dialogue sets with customer/agent labels, scenario metadata, intent categories, resolution outcomes, transcript structure, and QA fields for support automation evaluation.
Bilingual dialogue sets with language-span markers, speaker labels, timestamped transcripts, code-switch metadata, and aligned language fields for multilingual ASR and translation evaluation.
Fully scoped custom data collection with buyer-defined modality, schema, metadata fields, annotation layers, QA criteria, manifest structure, delivery package, and collection rules.
