Projection System

NeoArc's projection pipeline transforms a canonical domain model into purpose-specific representations for databases, search indexes, API contracts, UI wireframes, and generated documents.

April 1, 2026

From One Model to Every Target

Projections are the core mechanism in NeoArc Studio for turning a canonical domain model into the specific representations that downstream systems require. Each projection creates a derived entity graph with property-level mappings, type conversions, and transformation logic. The canonical model stays clean and focused on domain semantics, while projections handle the specifics of each target system - whether that is a relational database, a search index, an API contract, a UI wireframe, or a generated document.

Data Projections

Transform model entities for specific database targets with vendor-specific type resolution, naming conventions, and structural transformations.

Search Projections

Map model properties to search index fields with analysers, tokenisers, facet configuration, and vector embedding settings.

Schema Projections

Create type libraries for API contracts. Schema projections define reusable data structures that REST, GraphQL, gRPC, and other API types reference.

UI Projections

Build model-projected wireframes that bind UI layout and component placement directly to projected entity properties, giving end-to-end lineage from model to screen.

Document Projections

Define flow-graph pipelines that select, filter, loop, and format model data into structured documents with tables, charts, and responsive layouts.

Projection Sets

Group related projections as reusable templates with generic property slots. Define common patterns once and apply them across your workspace.

How Projections Work

The projection editor presents a mapper interface with source entities on the left and target entities on the right. Property-level mappings connect fields between the two sides. Each mapping can be a direct copy, a rename, a type cast, an aggregation, or a custom expression. The mapper tracks every connection so that lineage flows from the original model property through each transformation to the final target field.

Projection Chaining

Projections can chain: Model to Projection to Projection. Multi-layer architectures like medallion (bronze, silver, gold), data vault, or star schema are just chains of projections, each adding its own transformations. The first projection takes raw model entities and applies initial mapping. The second takes the output of the first and applies further refinement. Lineage flows through the entire chain, from the original model property to the final target field in the last projection.

Multi-Layer Architecture

Build medallion, data vault, star schema, or any layered data architecture by chaining projections. Each layer applies its own naming, typing, and structural rules.

Full Lineage

Trace any field in the final layer back through every intermediate projection to the original model property. Impact analysis follows the chain automatically.