On-Premise Deployment¶
Run NICL models on your own infrastructure for maximum control, security, and performance.
Why On-Premise?¶
Data Sovereignty |
Keep sensitive data within your infrastructure. No data leaves your network. |
Low Latency |
Eliminate network round-trips to cloud APIs. Ideal for real-time applications. |
Compliance |
Meet regulatory requirements (GDPR, HIPAA, financial regulations) with full control. |
Cost Predictability |
Fixed infrastructure costs instead of per-prediction pricing at scale. |
Offline Operation |
Run predictions without internet connectivity. |
Who Uses On-Premise?¶
Financial institutions with strict data residency requirements
Healthcare organizations handling patient data
Manufacturing with air-gapped production environments
Government agencies with security clearance constraints
Same SDK, Different Endpoint¶
Switching from cloud to on-premise requires just one parameter change:
# Cloud API
clf = NICLClassifier(api_key="nk_live_xxx")
# On-Premise - same interface
clf = NICLClassifier(host="http://your-server:8000")
Your existing code, pipelines, and workflows work unchanged.
Get Started¶
On-premise deployment is available for enterprise customers. Our team will help you with:
Infrastructure requirements and sizing
Deployment options (Docker, Kubernetes, bare metal)
Integration with your existing ML infrastructure
Ongoing support and updates
Contact us to discuss your requirements.