oruk: humanizing speech AI

Oruk provides English speech transcription, expressed-emotion and speaking-style labels, time-local segments, and optional speaker diarization through an API. Python and TypeScript SDKs and a hosted MCP server are available.

Key Features

Vocal Emotion and Delivery Analysis: Detects how words are spoken, identifying emotions such as frustration, excitement, hesitation, and sarcasm.
Resonance Model: Analyzes complete English recordings to return transcripts, emotion labels, speaking-style labels, and timed segments.
Realtime Model (Preview): Offers live multilingual transcription with phrase-level emotion across 32 locales with automatic language detection.
Broad Format Support: Processes audio files in WAV, FLAC, MP3, M4A, OGG, and WebM formats.
Data Privacy: Discards audio and model outputs after the response by default, ensuring customer data is not used for training without explicit written agreement.

Use Cases

Contact Centers: Review support calls with tone and context.
Voice Agents: Enable agents to pick up on hesitation, frustration, or excitement to shape their responses.
Sales Calls: Review conversations and coach sales teams.
Qualitative Research: Explore interviews with acoustic context.
Meetings: Revisit what was said and how it sounded.

Pricing

Plans start at $5 per month and include an audio allowance, with a 7-day free trial available to evaluate the API.

Getting Started

Try the public demo without an account; API plan and trial terms are on the pricing page.

https://oruk.ai/docs

https://oruk.ai/pricing

oruk

About oruk