errors with extra metadata. * Transcript validation – Whisper often hallucinates. We validate transcripts with our Olign tool, which provides reliable word- and utterance-level confidence scores. * Enriched labels – They do not include speaker names or talk-turns. We provide both. * Original data – They give you segmented audio only. We deliver full recordings with precise timestamps, and metadata, giving you more flexibility. * Customized Services – They leave you on your own. We provide tailored data processing services.

We released a small portion of tier IV of Olewave datasets for researchers to train their conversation voice AI models for free: https://huggingface.co/datasets/olewave/OleSpeech-IV-2025-EN-AR-100

The whole datasets are available for only US and Japan entities to purchase. We are a California company with no satellite office in China and no plans to establish one.

2. Voice data processing pipeline

Olewave’s innovative pipeline takes raw speech recordings, corresponding transcripts, and optional metadata as inputs, and robustly cleans noisy transcriptions and labels. It delivers validated speaker labels and transcripts, and rich features such as reliable word-level timestamps and confidence scores. By leveraging advanced alignment techniques, our pipeline ensures the highest levels of accuracy and usability, making it an indispensable tool for building high-performance but cost-efficient speech-based systems.

This pipeline or the modules in the pipeline are licensable.

3. Voice AI services

We build avant-garde but cost-effective in-house voice AI solutions such as ASR, TTS, speaker diariazation, conversational speech models ... in multiple languages by using our signature voice datasets and processing our client's voice data with our data processing pipeline. We have 20 years of experience in voice processing and 10 years of experience in tech consulting. We strictly adhere to our NDA, which is why client names are not shown on our website: https://www.olewave.com/