SPEECH DATA FOR AI · MAGHREB FIRST

Arabic dialect speech data, recorded where each dialect is spoken.

بيانات صوتية باللهجات العربية

Consented, native-transcribed speech in Tunisian, Libyan and Algerian Arabic — including the natural French–Arabic code-switching real people use — for training and evaluating speech recognition, translation and voice assistants.

✓ Native speakers ✓ Quality controlled ✓ Consent-based data
VOICE DATA DEMONSTRATION

Tunisian Arabic Speech

SAMPLE
TUNISIAN ARABIC ARABIC · FRENCH MIX
✓

Illustrative record: one audio clip, one metadata line. Real samples are shared under a non-disclosure agreement (NDA).

REAL DIALECTS. REAL VOICES. REAL DATA.
  • EU company · GDPR by default
  • Written consent from every speaker
  • Adults only, paid participation
  • Native-speaker transcription and review
  • Encrypted at rest and in transit
  • Rejected files replaced at our cost

Why now

Dialects are where speech models still fail.

Models handle Modern Standard Arabic well. Everyday speech in the Maghreb is a different story: dialect vocabulary, no fixed spelling, and constant switching into French. Closing that gap takes real recordings from real speakers, with rights cleared for commercial training.

~36%

average word error rate for the best systems on multi-dialect Arabic speech.

Source: NADI 2025 shared task (arXiv:2509.02038)

36.9%

word error rate on North African Arabic for the strongest open Arabic model tested.

Source: Cohere Transcribe Arabic release, July 2026

Dialect coverage
Maghrebi dialects are our core strength.
Tunisian
ar-TN
Pilot in preparation
Libyan
ar-LY
Next · on request
Algerian
ar-DZ
Phase 2 · on request
Moroccan
ar-MA
On request
Egyptian
ar-EG
On request
Levantine
ar-LB · SY · JO · PS
On request
Gulf
ar-AE · SA · KW · QA
On request
Modern Standard Arabic
MSA
On request
French–Arabic and English–Arabic code-switching is captured as it is spoken and marked in the transcript. Our Tunisian pilot dataset is being prepared now; other dialects are collected on request, to your specification.
Vision & Mandate
Make speech AI understand Arabic as it is really spoken.
Vision
Speech technology that works as well for a speaker in Sfax, Tripoli or Oran as for a speaker of Modern Standard Arabic, and later across the Arab world and Africa.
Mission
Collect consented, high-quality dialect speech on the ground, transcribe it with native linguists, and license it to AI developers with clear rights.
Mandate
Our own field operations in each country, consent and GDPR by default, original data only, and datasets delivered on time to the client's specification.
Where we are going
Now
Tunisia
Pilot dataset of paired conversation with French code-switching.
Next
Libya & Algeria
On request first, then our own datasets.
Then
Morocco & Egypt
Large dialect markets, opened with local coordinators.
Later
Gulf & Levant
Projects led from our Dubai base.
Later
West Africa
French–Wolof, Bambara and Dioula code-switching, built on our African networks.
Each new country opens only with a local coordinator, a check of local data law and a client need.
Services
Four services, one Arabic team.
Speech data collection
Scripted, prompted and conversational speech, recorded in-country against your quotas.
Transcription & annotation
Dialect-native orthographic and time-aligned transcripts, with speaker, noise and code-switching tags.
Arabic language services
Translation and translation verification, script verification, error creation and AI output evaluation.
Image data
Image collection and annotation, including Arabic text in images, on request.
Custom scenarios: in-car, call-centre, retail, smart-glasses and wearable commands; far-field or close-talk; real-world noise.
How we work
From your spec to delivered data.
01
Specify
Dialects, quotas, scripts, specs and acceptance criteria.
02
Recruit
Local coordinators recruit native speakers; regional origin checked at screening.
03
Record
Scripted, prompted or conversational, on smartphone, studio or wearable.
04
Review
Native review of every file, then dialect-native transcription.
05
Deliver
Weekly batches by secure SFTP or your cloud bucket.
Quality control. Rejected files are replaced at our cost.
We agree acceptance criteria with you — such as a WER target and a maximum rejection rate — and report against them for every batch.
Audio checks
Clipping, background noise, silence, overlap, duplicates.
Native review
Prompt adherence and dialect confirmation by a native reviewer.
Second-pass transcription
An independent reviewer checks a sample of each batch.
Batch reporting
Hours by dialect, city, gender, age band and rejection rate.
Responsibly sourced & Clearly documented
Every contributor signs a release for commercial AI training.
Written consent
Covers recording, transfer of rights and AI/ML training. Arabic and English versions. Voice-synthesis use is a separate opt-in.
GDPR by default
Pseudonymised speaker IDs. Names and signed forms stored apart from audio and never delivered. Consent records available for audit.
Ethical sourcing
Adults only, recruited directly — never through an anonymous crowd — and paid fairly and on time.
Data protection & security
Technical specs
Delivered in your format.
Capture
WAV, 48 kHz, 32-bit float, one lavalier track per speaker, backup recorder in parallel
Delivered audio
WAV, 16-bit PCM, mono · 16 kHz (44.1 / 48 kHz for clean or TTS data · 8 kHz for telephony)
Conditions
About 80% treated quiet room, 20% controlled real-world noise (café, car, office), each file tagged · or as you specify
Transcripts
Verbatim dialect in Arabic script, French/English in Latin script, 5–20 s time-aligned segments, event and code-switching tags
Metadata
JSON or CSV: speaker ID, gender, age band, dialect, region, environment, duration, QA status, consent scope
Delivery
Secure SFTP or your bucket (S3, Azure, GCS), weekly batches · we can adopt your schema and naming
Full technical specifications
Start with a pilot. Pay once you have validated it.
Step 1
Free sample
30 minutes of Tunisian speech with manifest and speaker sheet, to test against your spec.
No charge · under NDA
Step 2
Pilot
About 10 hours, one dialect, 20–30 speakers. Starts within 2 weeks of agreed specs; delivered in 3–4 weeks. No minimum order.
Fixed price, quoted per spec
Step 3
Production
Volumes scaled to your needs after the pilot, balanced by gender, age and city.
Per accepted audio hour
Custom
To your spec
Your prompts, domain, devices and noise conditions, on your platform or ours.
Per produced hour
Licensing: exclusive with full IP transfer, time-limited exclusive, or non-exclusive. Volume orders invoiced per accepted batch.

LEADERSHIP

Project discipline, technical delivery and
native linguistics.

MB

Mourad Ben Chedli

Founder · Project lead

PMP-certified, with a Master's in Project Management (Université du Québec en Outaouais). Brings 20 years of experience in sales, projects and operations across Africa, the Middle East, Europe and Canada, including IT projects for the Government of Canada.

IG

Ichrak Gharbi

Linguistic lead

Arabic language teacher and graduate of the University of Manouba (Arabic language and literature). Owns the transcription guidelines, trains transcribers and reviewers, and leads dialect quality control.

BM

Barhoumi Mustapha

Chief Technical Officer (CTO)

Web developer experienced in PrestaShop, PHP, JavaScript, HTML/CSS and API integration. His background includes e-commerce development, product-data automation and scraping, performance optimization, SEO, WordPress and database-backed applications.

AK

Dr. Achraf Kamoun

Head of Sound Engineering

Sound engineer with expertise in recording, acoustics and audio production. PhD in Cultural Sciences and Assistant Professor at the University of Tunis, specializing in fine arts. Founding member of the National Association of Music Therapy and founder of the Arab Board for Arts and Health. He holds several publications and patents in assistive technologies and has extensive experience organizing national and international conferences and festivals.

More about us and how we work →

Do you speak a Maghrebi dialect? Get paid to record
your voice.

شارك بصوتك باللهجة وخذ مقابل

Let's scope your Arabic pilot.

Tell us the dialects, hours and use case. We reply within one business day. Samples and the full specification are shared under NDA.

Request a sample

Phone

+216 23 22 22 66 (Tunisia)

+971 50 337 1701 (UAE)

+33 6 68 02 80 77 (France)

Offices

Brussels · Tunis · Dubai