# Glint

> Glint is building the largest GDPR-native AI training data infrastructure in Europe. The European alternative to Scale AI. Custom datasets from verified European contributors, delivered in 7 to 14 days, not 8 to 12 weeks.

## Markdown Mirrors

Each public page is mirrored as a clean markdown file for AI consumption (LLMs, search engines, scrapers). Use these instead of the HTML versions when you need the structured content without scripts, navigation, or visual junk.

- Homepage: https://glintdata.io/index.md
- For Contributors: https://glintdata.io/contribute.md
- For Enterprises: https://glintdata.io/datasets.md
- Blog index: https://glintdata.io/blog.md
- Join the Waitlist: https://glintdata.io/signup.md
- Privacy Policy: https://glintdata.io/privacy.md
- Terms of Service: https://glintdata.io/terms.md
- Blog post: Europe's Physical World Data Problem: https://glintdata.io/blog/europe-physical-world-data-problem.md

## About

Glint is a French B2B infrastructure company. We collect, verify, and license AI training data through a network of paid European contributors.

- Sector: AI training data infrastructure (B2B)
- Stage: Pre-seed (2M EUR round in preparation, 2026)
- Operating entity: Glint SASU (France, RCS Niort 104 088 703), wholly owned by holding company Airault & Co (France)
- Founder: Enzo Airault
- Headquartered in France, operating across the European Union
- Live since 2026

Glint is the European, GDPR-native alternative to Scale AI, Toloka, Encord, and micro1.

## Services

Glint operates an on-demand data collection infrastructure for AI labs, frontier model builders, and enterprise AI teams.

Data modalities collected:

- Audio and voice (speech, narration, voice commands, ambient audio, regional European accents and dialects)
- Video (standard format, egocentric / POV, demonstrations, walkthroughs)
- Screen recordings with voice (agentic AI, screen plus voice, computer use training data)
- Photo and image (object photography, scenes, document scans, annotated visuals)
- Text, transcriptions, annotations, translations
- Specialized expert data (medical, legal, scientific, technical) collected by domain professionals

Engagement model:

- Custom data collection campaigns scoped to client specifications
- Verified contributor network across European Union member states
- Quality validation, identity verification, and rights clearance built in
- Datasets delivered in 7 to 14 days for standard campaigns

## Pricing

- Custom pricing per dataset campaign, scoped on modality, volume, languages, quality grade, and exclusivity window
- Pilot and proof-of-concept campaigns available for new clients
- Contributors are compensated per task; payouts processed in 2 to 7 days via Stripe (card or SEPA bank transfer)
- Detailed pricing available on request through LinkedIn or the contact form

## Locations

- Headquarters: France
- Service area: European Union (27 member states) and EFTA countries
- Data hosting: European data centers, never leaves the EU
- Languages prioritized: French, English, Spanish, German, Italian, Dutch, Portuguese, Polish, plus regional and minority European languages (Breton, Occitan, Basque, Catalan, Welsh, Irish Gaelic, Sami, and others on request)

## Contact

- Public site: https://glintdata.io
- Contributor registration: https://glintdata.io/signup
- LinkedIn: https://www.linkedin.com/company/glint-data
- X / Twitter: https://x.com/glintdata_ai
- Sales and B2B inquiries: hello@glintdata.io or sales@glintdata.io, or reach out via LinkedIn direct message

## Service Area

- Primary market: European Union (all 27 member states)
- Secondary market: EFTA countries (Switzerland, Norway, Iceland, Liechtenstein) and the United Kingdom
- Contributors recruited and identity-verified across all EU member states
- Data residency: 100 percent European; no extra-EU data transfer

## Key Facts

- Glint is building the largest GDPR-native AI training data infrastructure in Europe
- Designed for compliance with the EU General Data Protection Regulation (GDPR) and the EU AI Act
- Direct competitors: Scale AI, Toloka, Encord, micro1, DoorDash Tasks
- Glint is the only European-headquartered, GDPR-native player in this category
- Target clients: AI research labs, frontier foundation model builders, world model developers, enterprise AI teams in healthcare, automotive, robotics, legal tech, and education
- Modalities supported: audio, video, egocentric POV video, screen plus voice, photo, image, text annotations, expert specialized data
- Standard delivery: 7 to 14 days; legacy data labeling firms deliver in 8 to 12 weeks
- Contributor network across the European Union with regional and dialect coverage

## What Makes Us Different

1. GDPR-native architecture, not retrofitted compliance. Every data point is collected with explicit, revocable consent under European law from day one.
2. European-first contributor network, not global crowdsourcing. Regional accents, minority languages, and cultural context are prioritized rather than averaged out.
3. Speed: custom datasets delivered in 7 to 14 days. Legacy data labeling firms take 8 to 12 weeks for the same scope.
4. Rare modalities supported by default: egocentric video (POV, smart glasses, GoPro), screen plus voice recordings, expert specialized annotations, regional European dialects.
5. Rights cleared by default. Every dataset ships with full legal documentation and licensing chain, ready for commercial AI training and frontier model training.
6. Built for frontier labs. Pricing, scoping, and SLAs are designed for AI research teams and foundation model builders, not generic data brokerage.
7. Data sovereignty: 100 percent of data is collected, processed, and hosted in the European Union. No data crosses outside the EU.

## FAQ

**Q: How is Glint different from Scale AI?**
A: Scale AI is US-based, operates a global crowdsourcing marketplace, and is not GDPR-native. Glint is a French infrastructure, with a curated and identity-verified European contributor network, every consent is explicit and revocable, every dataset stays in the EU, and rare European modalities (regional dialects, egocentric video, expert annotations) are core offer rather than edge cases.

**Q: How fast can a custom dataset be delivered?**
A: Standard custom datasets are delivered in 7 to 14 days, depending on modality and volume. Legacy data labeling firms typically take 8 to 12 weeks for similar scope.

**Q: What types of data does Glint collect?**
A: Audio and voice including regional European dialects, standard video, egocentric POV video (smart glasses, GoPro, head-mounted cameras), screen recordings paired with voice for agentic AI training, photos, images, document scans, text annotations, translations, and specialized expert data (medical, legal, scientific, technical).

**Q: Is the data GDPR compliant?**
A: Yes. GDPR compliance is built into the architecture from day one. Every data point is collected with explicit, freely given, revocable consent. All data is hosted in European data centers and never crosses outside the EU. Glint is also designed to comply with the EU AI Act for high-risk AI training data and provides full legal documentation per dataset.

**Q: How are contributors paid for their data?**
A: Contributors are paid per completed task. Payouts are processed in 2 to 7 days via Stripe (card or SEPA bank transfer). Each contributor signs a clear contribution agreement and can revoke consent at any time. Standard EU tax residency rules apply.

**Q: How can I contribute and earn from my own data?**
A: Visit https://glintdata.io/signup and register on the contributor waitlist. The first 100 verified contributors get early access to paid campaigns. Based on your equipment and skills, you can contribute audio, video, photos, screen recordings, or specialized expert annotations, and you are compensated per task.

**Q: Who are Glint's typical clients?**
A: AI research labs, frontier foundation model builders (Mistral AI, OpenAI, Anthropic, DeepMind, Cohere class), world model developers, robotics companies, autonomous driving teams, healthcare AI teams, legal tech companies, and any enterprise AI team that needs compliant European training data at scale.

**Q: How do I request a custom dataset for my AI training?**
A: Reach out via LinkedIn at https://www.linkedin.com/company/glint-data, or use the contact form on https://glintdata.io. To accelerate scoping, include modality (audio, video, image, screen recording, text, expert), target volume, target languages, quality grade, and the AI use case the data will train.