Linux, cloud infrastructure, containers, supercomputing, Python, PyTorch, and much of today's AI ecosystem grew because people and organizations worked together in the open. OpenWALDO brings that proven community model to one of AI's most important foundations: the data, artifacts, licenses, and provenance behind the models.
OpenWALDO is an open source AI community building a shared public corpus and the tools needed to carry its history into trained model releases. Instead of every team repeatedly assembling the same foundational material, we can build one accountable commons together—then give every builder the freedom to select, combine, train, and innovate above it.
A source added once can support many models. A correction strengthens the public record. Better ingestion, review, and training tools help every future contributor. Individuals, hobbyists, researchers, institutions, and companies can learn together, contribute their expertise, and build from the same public foundation.
The commons does not prescribe one dataset or one model. Builders choose the material appropriate for their work and remain free to create, differentiate, and compete. Shared foundational work moves innovation higher up the stack.
OpenWALDO treats training data as inspectable infrastructure:
- Git governs meaning. Sources, license assertions, corpus structure, counts, and object references remain readable, reviewable, and attributable.
- Content hashes establish identity. Canonical data is stored by content, making exact inputs independently verifiable.
- AI Bills of Materials preserve the record. Corpus selections, licenses, starting weights, training runs, and release artifacts stay connected across the model lifecycle.
- DCO sign-off preserves accountability. Responsibility remains attached to contributions as the community and corpus grow.
- Federated storage supports public scale. Git carries the reviewable record while content-addressed lookaside storage carries the data.
The result is a durable public AI commons: community governed, independently verifiable, and open for everyone to improve, learn from, and build upon.
| Path | What you can do |
|---|---|
| Explore the public corpus | See how sources, licenses, provenance, and corpus Bills of Materials remain connected. |
| Build and train | Compose, verify, train, and export models without dropping their lineage. |
| Contribute to the corpus | Turn public source material or reviewed acquisition recipes into accountable corpus contributions. |
| Join the AI community | Learn together, get help, collaborate, and help lead the project. |
openwaldo/waldo— the command-line toolchain for indexing, verification, corpus export, training, provenance, and model release.openwaldo/waldo-index— the Git-governed public corpus index.openwaldo/openwaldo.org— the OpenWALDO website and project documentation.openwaldo/logos— distributable PNG and SVG brand assets.
Build the foundation together.
Website ·
Slack ·
Contributing