SureCentric Overview
Connect to, collect, analyze — any apps, any data, any AI.
SureCentric is an AI-native platform for centralized clinical intelligence purpose-built for regulated health sciences enterprises. It unifies three traditionally separate worlds — ontology-driven schema management, dynamic data warehousing, and natural-language AI analytics — into a single, auditable platform.
The Challenge
Clinical organizations operate across a fragmented landscape: SureDrive document archives, external databases, laboratory systems, EDC platforms, and cloud APIs — each with its own schema, its own format, and its own access model. Connecting them for reporting and analytics today means months of custom integration work, brittle ETL pipelines, and query results that cannot be reproduced or audited.
SureCentric solves this by putting schema first. Every data source, every query, every dashboard starts with a declared, versioned, semantic schema — and that schema drives everything downstream automatically.
Core Principles
| Principle | Meaning |
|---|---|
| Accurate | SQL generated from deterministic, declared schema — not hallucinated joins |
| Reliable | Same schema + same data + same query = same result, every time |
| Reproducible | Versioned schemas, immutable data snapshots, auditable lineage |
| Performant | Columnar execution (DuckDB/Arrow), async workers, cached compiled models |
| Portable | JSON-LD schemas are self-describing and vendor-neutral — export to OWL, DDL, or any target |
What SureCentric Does
SureCentric bridges three traditionally separate technical domains:
| Domain | SureCentric Approach |
|---|---|
| Knowledge Graphs & Ontologies | JSON-LD as the canonical semantic layer; OWL import/export; SPARQL autocompletion |
| Data Lakehouses | DuckDB / PostgreSQL for execution; Apache Iceberg for time-travel; Trino for federated SQL |
| AI-Native Operations | LiteLLM as the provider-agnostic AI gateway; NL2SQL via DataAgent; agentic workflows via SureAgent |
Platform Diagram
Standalone and Integrated
SureCentric is designed to operate in two modes:
Standalone Mode
SureCentric can be deployed as a self-contained platform — independent of any document management system. In this mode it connects directly to external databases, cloud data warehouses, REST APIs (via SureConnect), and file sources (Parquet, CSV, JSON-LD). This makes it suitable for:
- Analytics teams working outside a SureDrive environment
- Organizations integrating SureCentric with third-party platforms (Snowflake, BigQuery, Salesforce Health Cloud)
- Future integrations with any OpenAPI-compliant platform
Integrated Mode (SureClinical Platform)
When deployed alongside SureDrive, SureCentric sits above the document archive layer and federates queries across multiple SureDrive applications — each with its own content model — into a single unified lakehouse schema. Schema Builder acts as the bridge: it imports OWL content models from SureDrive archives and generates the SQL DDL that powers the analytics warehouse.
Platform Components
| Component | Role |
|---|---|
| Schema Builder | AI-assisted JSON-LD schema editor — the primary authoring tool for all schemas |
| SureSchema | Deterministic semantic model composed from ordered JSON-LD schema arrays |
| DataAgent | NL2SQL engine grounded by compiled schema metadata |
| SureAgent | Agentic workflow orchestration |
| SureLLM | Private domain LLM gateway (LiteLLM-compatible) |
| SureIntel | RAG pipeline (Flowise-based) for document intelligence |
| SureIndex | Universal data ingestion via CocoIndex |
| SureConnect | OpenAPI connector for any REST API source |