Implicit geometry for physical design and engineering.
NeuroCAD is building a production-oriented engineering stack across implicit geometry, solver-ready data generation, model training workflows, and enterprise integration surfaces.
This GitHub organization is the public technical layer around that system.
It exists to expose the parts of NeuroCAD that should be inspectable by partners, customers, and the engineering community: API contracts, benchmarks, examples, and shared technical documentation.
NeuroCAD already operates across four technical layers:
- implicit geometry infrastructure for physical design workflows
- simulation-grade data generation pipelines
- model training and evaluation workflows
- enterprise-facing integration surfaces
The public repositories in this organization are intended to show real technical structure and real engineering output without disclosing proprietary implementation details.
Current public repositories:
neurocad-api-docsfor API contracts, guides, and SDK-facing examplesneurocad-benchmarksfor selected benchmark artifacts, metadata, and protocol notesneurocad-examplesfor reference integration examplesneurocad-architecturefor public system overview and boundary documentation.githubfor organization profile and default governance files
The public layer is intentionally structured to show engineering discipline:
- versioned GitHub releases across the public repositories
- validation workflows on contracts, examples, benchmark packages, and architecture docs
- explicit public schemas and sample artifacts
- release notes and checksums for benchmark packaging
- architecture decision records for public boundary choices
This is not a marketing mirror. It is the inspectable technical shell around the private platform.
The primary public documentation landing page for integrations is:
The following layers remain proprietary:
- production kernel internals
- large-scale training infrastructure
- internal data generation pipelines
- deployment and operational systems
This is a deliberate boundary. We believe a serious engineering company should expose a clear technical surface to the world while protecting the implementation details that define its core IP.
NeuroCAD is built for industrial and enterprise contexts where geometry, simulation, data pipelines, and system integration have to be credible, reproducible, and operationally disciplined.
If you are evaluating NeuroCAD for:
- enterprise integration
- technical partnership
- research collaboration
- infrastructure or model diligence
please contact us directly.
- use GitHub issues for repository-scoped documentation or integration questions
- use private email for security disclosure and diligence paths
- expect the public surface to evolve through tagged releases rather than silent changes
- Website: neurocad.eu
- Contact: office@neurocad.eu