One Plugin.
Four AI Engines.
Zero Architecture Changes.
ONE Multi-Target Application deployed in 12 minutes. Modular licensing controls which engines are active. Disabled engines consume zero CPU.
ONE MTA vs. 4 Separate Plugins
Why deploy four separate applications when one unified architecture delivers more for less?
4x infrastructure cost. Zero cross-engine intelligence. 4 separate deployments to manage.
63% lower cost. Native cross-module intelligence. One deployment command.
6-Layer Architecture
From the user interface to your S/4HANA system — every layer is purpose-built, security-hardened, and optimized for CPU-only AI inference.
User Interface
Unified dashboard across all four engines with role-based views
Application Router
Single entry point with enterprise-grade security headers
Platform Service
Core runtime: all 4 engines + cross-module intelligence layer
Data Layer
Single HDI container with shared + domain-specific CDS entities
SAP Integration
Native BTP service bindings for event-driven architecture
Customer's S/4HANA
Read-only access via standard SAP APIs. No ABAP. No transports.
Zero Modifications to S/4HANA
All communication uses standard OData APIs. No ABAP code. No transports. No BAdIs. 100% Clean Core compliant.
Multiple Domains Flag the Same Entity?
Scores Multiply Automatically.
When multiple engines flag the same vendor, employee, or transaction, the correlation engine automatically boosts risk scores. Zero configuration required.
Registered domains are automatically discovered and included in correlation analysis.
Levenshtein distance of 3 or less matches entities across domains, catching typos and variations.
No setup, no rules, no thresholds. The correlation engine works the moment you deploy.
License JWT Controls Activation.
Disabled Engines = Zero CPU.
Pay only for the engines you use. Disabled engines are code-present but return immediately, consuming zero processing resources.
{ "apps_enabled": ["fraudguard"] }
{ "apps_enabled": ["fraudguard", "compliance", "supplychain", "voiceops"] }
Adding Domain 5 Requires
Only 5 Steps.
The architecture is built for extensibility. New AI engines plug into the existing platform without changing a single line of core code.
What You DO
Implements BaseEngine interface
1 line of code
Domain-specific tables
cf deploy zynoviq-sap-plugin.mtar
Enable new domain in JWT
What DOESN'T Change
The core platform is locked. New domains plug in. They never modify the foundation.
7 Modules. 5 BTP Resources. One File.
Everything your SAP system needs — from AI inference to auto-updates — compiled into a single deployable .mtar archive.
7 MTA Modules
All engines + AI models
HANA HDI schema deployment
Authentication + routing
Unified dashboard
License validation
Auto-updates
Joule skills/agents
5 BTP Resources
Persistent data layer
Authentication & authorization
Event-driven messaging
S/4HANA connectivity
On-premise tunnel
One Command Deployment
All 7 modules deployed, all 5 BTP service bindings created, all HANA tables provisioned. 12 minutes. One command.
CPU-Only Performance Targets
All benchmarks measured on standard BTP infrastructure: 8 vCPU, 32GB RAM. No GPU. No specialized hardware. Production-grade performance.
Architecture Your CIO
Will Love
One MTA. Four AI engines. CPU-only. Clean Core compliant. 12-minute deployment. 63% lower infrastructure cost. Zero architecture changes.