Content Safety Models That
Get Smarter Together

Enterprise content moderation. Train on your infrastructure, access foundational models trained across the industry, benefit from network effects as the cooperative grows. Patent protection included.

Patent Portfolio: WO2022137077A1 • Granted in EU (15 countries), UK, South Africa • US Pending
Cooperative Benefits: Foundational models • Federated learning • Expert ML support • 50% cost savings
Explore MembershipHow Cooperation Works

You Need Patent Licensing. Here's The Smart Way

Modern content moderation systems use AI-powered inspection, user-specific restrictions, and real-time decisions. These architectures may practice our granted patents (WO2022137077A1, granted in EU, UK, South Africa; US pending).

If you're running content moderation at scale, you likely need a patent license. We offer two paths forward.

What Patents Cover
  • ✓ Dual-factor AI moderation (rating + confidence scoring)
  • ✓ Real-time visual content inspection
  • ✓ User-specific restriction parameters
  • ✓ Selective content redaction based on AI analysis
Patent Status
  • 15+ jurisdictions granted
  • EU, UK, South Africa
  • US Patent Pending
  • Priority Date: December 2020

Two Ways to License

Both options provide full patent protection. Cooperative membership adds foundational models, infrastructure flexibility, and expert access—at 50% lower cost.

Standalone License

£25K/year

starting at

scales to £1M+ for enterprise

What's Included:

  • Patent license protection
  • Compliance certificate

What's NOT Included:

  • Models or AI infrastructure
  • Technical support
  • Ongoing improvements
  • Cooperative benefits

You'll Also Need:

  • Dedicated ML engineering team
  • Infrastructure & cloud services
  • Continuous R&D investment
Request License Quote
⭐ RECOMMENDED

Cooperative Membership

£12K/year

starting at

scales to £500K+ for enterprise

50% savings at every tier

Everything Included:

  • Patent license + compliance
  • Access to foundational models
  • Federated learning infrastructure
  • Monthly model updates
  • Integration support
  • Network effects benefits

How Training Works:

  • Train on your servers with your data
  • Share only encrypted model weights
  • Your data never leaves your infrastructure
  • You own your models - no lock-in
From £12K/year

Founding member rate - locked in for 3 years

Join Cooperative

Both options provide patent licensing. Cooperative membership adds model access, infrastructure, and ongoing support—at 50% lower cost than standalone licensing alone.

How Cooperation Creates Value for Everyone

Your participation improves models for the entire network—including yours. Cooperative economics mean better models at lower cost than any platform could achieve alone.

How It Works

1

You Train Locally

Use your own data on your servers. Your content never leaves your infrastructure.

2

Share Encrypted Weights

Contribute model parameters only (not data). Differential privacy protects individual examples.

3

We Aggregate

Combine weights from all members using secure aggregation protocols.

4

Everyone Benefits

Receive improved models trained across diverse data you couldn't access alone.

The Cooperative Flywheel

Hyperscalers Contribute

Global data diversity, billions of edge cases, multi-language training, regulatory compliance experience.

→ Improves foundational models for everyone

Regional Platforms Add

Local context, cultural nuances, language-specific patterns, niche content types.

→ Models work better across cultures

Our Expert Team Provides

Specialized ML expertise, custom model training, regulatory compliance guidance, continuous optimization.

→ Every member benefits from expert support

Everyone Benefits

Better accuracy, comprehensive coverage, regulatory compliance, continuous innovation, lower costs.

→ Industry-wide improvement

No Single Platform Could Build This Alone

Hyperscalers need global diversity. Regional platforms need scale. Startups need accessibility. Model trainers need data. Together, we create industry-leading models that serve everyone.

Clear, Predictable Costs

Transparent annual pricing from startups to enterprise platforms. All memberships include patent licensing, foundational model access, and cooperative participation. Founding members get locked in rates.

⭐ Founding Member Pricing

First 20 members get exclusive rates: £12K-£500K/year locked in. After 20 members, standard rates apply. This is a limited-time opportunity to establish founding member economics.

Platform SizeStandalone LicenseCooperative MemberAnnual Savings
Startup
<1M users
£25K£12K£13K (50%)
Growth
1-10M users
£100K£50K£50K (50%)
Mid-Market
10-100M users
£400K£200K£200K (50%)
Enterprise
100-500M users
£1M£500K£500K (50%)

Platforms serving 500M+ users: Custom enterprise agreements available. Contact us for board-level negotiations including governance participation and multi-jurisdiction coverage.

Pricing Based On

Annual content volume (images, videos, text), active users, API request volume. Custom pricing available for unique situations.

What's Included

Full patent license, federated learning infrastructure, foundational models, monthly updates, integration support, compliance docs, governance (Enterprise+).

Volume Discounts

Multi-year commitments, early adopter rates for founding members, enterprise custom agreements.

Get Custom Quote

Why Every Platform Benefits

The cooperative creates value across the entire industry. Network effects improve models for everyone—from startups to hyperscalers.

Startups
Enterprise-grade models at £12K/year. Access technology used by platforms 1000x your size. Compete on features, not infrastructure.
Mid-Market
Diverse training data improves edge cases. Custom solutions from expert trainers. £200K vs £2M+ building alone.
Hyperscalers
Global data diversity improves models. Regulatory compliance certified. Patent clarity. Your participation strengthens the entire network.

Building the Network

20+
Platform target in 18 months
Onboarding
Founding members across all tiers
All Scales
From startups to hyperscalers welcome

As the Network Grows

5
Members
Basic model coverage, limited diversity, foundation established
10
Members
Improved accuracy, better edge cases, regional coverage
20+
Members
Industry-leading models, comprehensive coverage, continuous innovation

Why Federated Learning Matters

Traditional Centralized Training

  • ✗ All partners send raw data to central server
  • ✗ Massive GDPR and privacy concerns
  • ✗ Competitive data exposure
  • ✗ Single point of security failure
  • ✗ Platforms will never participate

Our Federated Approach

  • ✓ Train on your servers with your data
  • ✓ Share only encrypted model weights
  • ✓ No competitive exposure
  • ✓ GDPR compliant by design
  • ✓ Better models through collaboration

Enterprise-Grade Infrastructure

Our federated learning platform provides production-ready infrastructure for platforms at any scale.

Privacy & Security

Differential Privacy

ε=1.0, δ=10⁻⁵ guarantees protect individual examples

Secure Aggregation

Encrypted weight transmission using secure multi-party computation

Compliance

SOC 2 Type II certified • GDPR compliant by design • DSA/OSA ready

Integration

APIs & SDKs

REST + gRPC APIs • Python, JavaScript, Java, Go SDKs

Deployment

Docker containers • Kubernetes-ready • Average integration: 2-4 weeks

Documentation

Complete integration guides • Sample code • Technical support

Model Performance

Latency

<30ms inference • <10ms for cached results

Availability

99.9% uptime SLA • Global edge deployment

Scale

Handles Billions of requests/day • Auto-scaling infrastructure

Accuracy

92%+ on standard benchmarks • Improves with network growth

Download Technical WhitepaperView API Documentation

Technical Deep Dive Available: Schedule a discovery call for detailed architecture review, security audit results, and integration planning with our engineering team.

Simple Onboarding

From first call to production integration in 4-6 weeks. We make joining the cooperative straightforward.

1
Discovery Call

30 minutes

  • • Discuss your moderation needs
  • • Review current implementation
  • • Assess patent licensing requirements
  • • Answer technical questions
2
Legal Review

1-2 weeks

  • • Sign membership agreement
  • • Finalize patent licensing terms
  • • Define integration scope
  • • Establish governance rights
3
Technical Integration

2-4 weeks

  • • Integrate federated learning SDK
  • • Connect to model serving
  • • Train initial models on your data
  • • Go live with production traffic

Ready to discuss your platform's needs?

Common Questions

Clear answers to help you understand our cooperative model and patent licensing.

What exactly does the patent cover?

Our patents cover AI-based content moderation systems that use dual-factor decisions (rating + confidence level), real-time visual content inspection, user-specific restriction parameters, and selective content redaction. See our patent documentation for detailed claims.

How do I know if I need a license?

If you're using AI to inspect content in real-time, make moderation decisions based on multiple factors, and apply user-specific restrictions, you likely practice our patent claims. We can review your implementation during a discovery call.

Why is cooperative membership cheaper than standalone licensing?

Members contribute to federated learning, creating value for the entire network. We share those economics through reduced rates. You also avoid building infrastructure yourself.

What data do I have to share?

You share encrypted model weights only—never raw data, content, or competitive information. Your training data stays on your servers. Federated learning means privacy by design.

What if I want to leave the cooperative?

Members can exit. You keep your trained models but lose access to future updates and must negotiate standalone patent licensing.

Do I have to contribute to federated learning?

Active participation (sharing weights) is required for cooperative membership. If you want patent license only, choose standalone licensing option.

How many members are there currently?

We're onboarding founding members across all tiers—from startups to hyperscalers. First 20 members get locked-in pricing and governance seats. Target: 20+ platforms in 18 months to achieve network effects.

Do I train on my own infrastructure?

Yes. Federated learning means you train locally on your servers—your data never leaves your infrastructure. Only encrypted model weights are shared with the cooperative. This is how we maintain privacy while enabling collective intelligence. We provide SDKs, integration support, and initial model weights to get you started.

Can I see the code for federated learning infrastructure?

We provide technical documentation and integration specs. Core infrastructure is proprietary but auditable for security/privacy verification.

Is this compliant with EU regulations?

Yes. Federated learning is GDPR-compliant by design (no data transfer). Our architecture supports DSA and Online Safety Act requirements. We provide compliance documentation.

Still have questions?

Schedule a 30-minute discovery call to discuss your specific situation.

Schedule Discovery Call