Platform Foundation

SureCentric is built on a layered architecture that cleanly separates data sources, semantic schema, AI services, and analytics — allowing each layer to be upgraded or replaced independently. This page describes the foundation layer: how the platform is structured, how it can be deployed, and how components integrate.

Layer Model

┌─────────────────────────────────────────────────────────────────────┐
│                   SureCentric Desktop                               │
│   Import, View, Report, Query, Explore                              │
│   ┌────────────┐ ┌────────────┐ ┌───────────┐ ┌───────────────────┐│
│   │ Dashboards │ │ Query &    │ │ Explorer  │ │ Schema Builder    ││
│   │ (Superset) │ │ Reports    │ │ & Viewer  │ │ (JSON-LD Editor)  ││
│   └────────────┘ └────────────┘ └───────────┘ └───────────────────┘│
├─────────────────────────────────────────────────────────────────────┤
│         Dynamic Data Warehouse / Lakehouse                          │
│         (DuckDB / PostgreSQL / Apache Iceberg)                      │
│  ┌──────────────────────────────────────────────────────────────┐   │
│  │  Dynamic Schema — auto-generated from JSON-LD                │   │
│  │  CREATE TABLE · VIEW · INDEX · FK constraints · CHECK        │   │
│  └──────────────────────────────────────────────────────────────┘   │
├─────────────────────────────────────────────────────────────────────┤
│  Schema Layer (JSON-LD Canonical)                                    │
│  Schema Builder → Mappers/Translators → OWL · DDL · JSON Schema     │
├─────────────────────────────────────────────────────────────────────┤
│  AI Layer                                                            │
│  SureAgent · SureIntel (RAG) · SureLLM · SureIndex                  │
├─────────────────────────────────────────────────────────────────────┤
│  Data Sources                                                        │
│  SureDrive Apps · External DBs (PG, Snowflake) · SureConnect (API)  │
└─────────────────────────────────────────────────────────────────────┘

Deployment Modes

Standalone Deployment

SureCentric can run entirely independently of SureDrive or any SureClinical application. In standalone mode, the platform connects directly to:

• External SQL databases (PostgreSQL, MySQL, Snowflake, BigQuery)
• Data lakehouses (Apache Iceberg, Delta Lake, Parquet files)
• REST APIs via SureConnect (any OpenAPI 3.x-compliant service)
• File sources (CSV, JSON, JSON-LD, XLSX)

This deployment model is ideal for:

• Analytics and data engineering teams at organizations that do not use SureDrive
• Integrations with third-party clinical platforms (Veeva Vault, Medidata, Oracle Health)
• Future interoperability with any platform that exposes structured data via API

The Schema Builder remains the primary authoring tool in standalone mode — schemas are authored against whatever data sources are connected, and the DDL pipeline generates the lakehouse tables automatically.

Integrated Deployment (SureClinical Platform)

When deployed alongside SureDrive, SureCentric federates queries across multiple SureDrive applications — each with its own OWL-based content model — into a unified lakehouse schema.

SureDrive App1 (CM1) ──┐
SureDrive App2 (CM2) ──┼──► SureArchive (OWL aggregate) ──► Schema Builder ──► Lakehouse DDL
SureDrive App3 (CM3) ──┘

Schema Builder imports these OWL content models, merges them into a unified JSON-LD schema, and generates the PostgreSQL/DuckDB DDL that powers Superset dashboards.

Core Infrastructure

Data Warehouse

Engine	Role
DuckDB	In-process analytical engine; fast local execution; Arrow-native
PostgreSQL	Persistent relational store; CARD plugin profiles; schema metadata
Apache Iceberg	Time-travel and schema evolution for large-scale lakehouse tables
Trino	Federated SQL across enterprise data sources

Schema Pipeline

The schema pipeline is the foundation of SureCentric's accuracy guarantee:

JSON-LD Schema (authored in Schema Builder)
        │
        ▼
Schema Translation Service
        │
        ├──► SQL DDL (PostgreSQL / DuckDB CREATE TABLE)
        ├──► OWL / RDF-XML (Nuxeo archive format)
        └──► JSON Schema (API validation / registries)

Because SQL is generated from a declared, versioned schema — never inferred from physical data — query results are deterministic and reproducible across environments.

Key Concepts

Concept	Description
JSON-LD	W3C JSON-LD 1.1 — the canonical schema format and single source of truth for all authoring
SureSchema	Deterministic semantic model composed from an ordered array of JSON-LD component schemas
Plugin System	ES-module plugins extend Schema Builder with custom translators, panels, importers, and validators
DDL Pipeline	JSON-LD → Mapper View → Plugin Translator → CREATE TABLE statements
CARD Profile	Named, persisted configuration that captures schemas, datasets, plugins, and BI tool targets
OWL Sidecar	Spring Boot service for OWL/RDF-XML import/export (SureDrive archive integration)