Introdução

Many biometric projects begin with a simple goal:

• launch an MVP
• validate the business model
• test real-world deployment

However, as the project grows, a major concern appears:

👉 How can the system scale to millions of users without disrupting existing customers?

BioWavePass was designed with this challenge in mind from the beginning.

Its palm vein architecture supports a smooth transition from Small Model deployment to Large Scale Model infrastructure while keeping the user experience unchanged.

Built for Growth from Day One

Traditional biometric systems often create problems when scaling:

• database migration becomes complex
• users must re-register
• infrastructure requires rebuilding
• customer experience is interrupted

BioWavePass uses a different architecture.

The system is designed so customers can:

✅ Start small
✅ Expand gradually
✅ Upgrade smoothly

without rebuilding the entire platform.

Small Model: Fast and Lightweight Deployment

The Small Model is designed for:

• MVP projects
• Proof of Concept testing
• pilot deployment

It supports:

✅ Up to 10,000 free user IDs

The Small Model uses only two device-side biometric features:

1. Palm Print Feature (RGB)
2. Palm Vein Feature (IR)

Only feature vectors are uploaded by the SDK.
RGB + IR images are stored separately as backup.

Why the Large Scale Model Matters

As biometric systems grow larger, scalability becomes more important.

The Large Scale Model introduces:

• server-side verification
• GPU-based processing
• scalable vector database architecture

allowing the platform to support:

✅ 100K to millions of users

while maintaining:

• ~0.35s matching speed
• ~99.8% accuracy

The Key to Seamless Migration

The most important design advantage is this:

👉 Existing RGB + IR registration images can be reused during migration.

This means users originally registered in the Small Model system do not need to scan their palm again.

How the Migration Process Works

From the application perspective:

• the system invokes the Large Model registration interface
• stored RGB + IR images are programmatically registered into the new system
• the Large Model rebuilds the biometric registration internally

From the user perspective:

👉 Nothing changes.

Users do not need to:

• re-register
• capture new images
• repeat onboarding

The migration is effectively invisible to the end user.

Why RGB + IR Image Storage Is Important

Image storage is the foundation of future scalability.

In the Small Model:

• daily matching uses feature vectors only
• images are backup only

In the Large Model:

• the algorithm server re-extracts features from stored images
• server-side verification becomes part of the matching process

This creates a more scalable biometric architecture.

Large Model Verification Architecture

The Large Scale Model uses four verification elements:

1. End-side palm print feature
2. End-side palm vein feature
3. Server-side palm print feature from RGB image
4. Server-side palm vein feature from IR image

This dual-layer architecture improves:

• database scalability
• verification stability
• large-scale matching performance

Lower Operational Risk

BioWavePass helps reduce long-term deployment risk because customers can:

• launch quickly with Small Model
• validate market demand
• scale only when needed

without forcing users through another registration process.

Development and Testing Support

After sample purchase, BioWavePass provides:

• Small Model test server
• Large Scale Model test server
• API integration support
• debugging environment

This allows customers to verify the migration path during development.

Customer-Controlled Infrastructure

All systems are deployed on customer-owned servers.

This ensures:

• customer-controlled biometric data
• GDPR-compliant architecture
• no vendor-side biometric data management

BioWavePass also provides:

✅ Up to 5 years large-scale algorithm support

Conclusão

A scalable biometric platform should support long-term growth without disrupting users.

BioWavePass enables organizations to:

• start with MVP deployment
• migrate programmatically
• scale to millions of users

while keeping the transition seamless for end users.

Final Thought

The best biometric migration process is the one users never notice.

CTA

Learn more about BioWavePass Palm Vein Technology:

👉 https://biowavepass.com/biowavepass-palm-vein-scanner-products/