Voice Harbor by Nijta

Voice Harbor removes PHI & biometric voiceprints from your voice data that helps you unlock its full access, extract insights and drive innovation - while being 100% compliant.

Book a discovery call

API Documentation

remove the id. keep the insights

Advanced
speech-to-speech
anonymisation

biometric anonymisation

Voice identity removed

Through its proprietary gen-AI speech-to-speech solution, Nijta transforms voice data in a way that the speaker's identity becomes fully untraceable.

Customised solution

Tailored to your use case

Nijta adapts to the target geographical, sociocultural conditions and industry-specific PHI and confidential information of your data.

Sensitive content redaction

PHI and confidential data accurately filtered out

Unlock the potential of your voice data with an automated high-fidelity anonymisation process. Start innovating, experimenting and growing faster.

superior fidelity

Retain all emotions & data value

Content, prosody and emotions are key to teaching your AI models and mining your data. With Nijta, your data stays 100% useful and natural, so no insight is left behind.

Zero PHI left

No corruption of your data

Legally guaranteed

Validated by the French Data Protection Agency (CNIL)

Irreversible anonymisation

No way back to the speaker

Clean your data, in seconds

Python


    from client import VoiceHarborClient
    
    BASE_URL = "https://voiceharbor.ai"
    admin_token = <Contact sales to receive your ADMIN TOKEN>
    usage_token = VoiceHarborClient.create_developer_token(BASE_URL, admin_token)
    job_id = VoiceHarborClient.create_job(BASE_URL, usage_token)
    
    client = VoiceHarborClient(
        base_url=BASE_URL,
        job_id=job_id,
        token=usage_token,
        inputs_dir="<DIRECTORY_TO_AUDIO_FILES>",
    )
    
    job_params = {
        "agents": ["health-generic"],
        "model": "advanced",
        "files": []
    }
    
    job_params = client.submit_files(job_params)
    downloaded_results = client.download_results(output_dir="./results")

Book a discovery call

PHI Redaction

Automatically redact PHI from voice data

Supported in 90+ languages

Ensure compliance with GDPR, HIPAA & AI Act.

Biometric Anonymisation

Automatically remove biometric identifiers irreversably from voice data to ensure protection of speaker's identity.
Supported in French & English, more languages coming soon.

Speaker Diarization

Identify, segment, and separate multiple speakers conversing in a voice sample for efficient analysis.

Speech-to-Text

Convert audio and video files, as well as live speech into text with precision.

Set your data free
Start using Nijta.

Anonymise your first voice recording in minutes.

Watch a demo