# STRUC | The Document Infrastructure Company

## Mission
STRUC builds high-performance C++ document infrastructure. We provide the missing last mile for AI: a foundry for generation and an index for retrieval.

## Core Products

### 1. Foundry (Generation Engine)
A C++ headless engine that exposes a direct object graph to write, edit, and mutate high-fidelity documents via atomic transactions.
- **Performance**: Engineered for <50ms latency. Zero-serialization overhead.
- **Interface**: JSON-RPC. Direct Object Graph Injection.
- **Reliability**: Strict schema enforcement prevents LLM hallucination of formatting.
- **Roundtripping**: High-fidelity DOCX -> Object -> DOCX cycles preserving comments, track changes, and margins.
- **Output**: Universal output targets (DOCX, PDF/A).

### 2. Index (Retrieval Engine)
Structure-Aware RAG (Retrieval-Augmented Generation).
- **Problem**: Standard vector databases slice documents blindly, destroying context.
- **Solution**: STRUC Index exports Semantic Chunks that respect semantic boundaries for perfect retrieval.
- **Result**: Atomic Objects. Perfect Retrieval.

## Use Cases & Industries
- **LegalTech**: Deterministic redlining, strict version control.
- **FinTech**: High-volume automated reporting, audit trails.
- **Healthcare**: Tamper-proof, audit-ready patient records.
- **GovTech**: ISO-compliant archival (PDF/A).

## Technical Specifications
- **Latency**: 42ms (Benchmark: Native C++ Engine vs 2400ms Python/XML Parsing).
- **Compliance**: PDF/A-1b native generation.
- **Integration**: Universal Agent Interface (supports OpenAI, Anthropic, Llama).

## Access
Status: Private Beta. Invite Only.
To request API keys, visit https://struc.so and join the waitlist.
We are manually onboarding high-volume partners.