[feat]: Improve developer onboarding and example integrations

[0xRejik]

Product

SDK (JS/TS)

Problem to Solve

Many developers who discover Shelby are interested in using it as a data layer for AI applications, but the onboarding process can still feel a bit unclear for first-time users.
Currently, the documentation explains the concepts well, but practical examples for real-world use cases are still limited. Developers may struggle to understand how Shelby integrates into existing AI pipelines such as:
storing training datasets
retrieving structured data for inference
building verifiable storage for AI agents
Without clear end-to-end examples, new builders may take longer to experiment or may not fully realize Shelby’s potential as a global object storage layer designed for AI.

Proposed Solution

Add step-by-step quickstart tutorials that guide developers from installation to a working example

Provide practical sample projects, such as:
AI dataset storage with Shelby
Retrieval of verifiable data for AI agents
Integration with common AI frameworks

Create template repositories showing Shelby integration with modern stacks (Node.js / Python)

Include architecture diagrams that illustrate how Shelby fits into an AI data pipeline

Provide a minimal example SDK workflow that demonstrates:
upload object
verify object
retrieve object in an AI workflow.

Alternatives Considered

Shelby has strong potential as a verifiable storage layer for AI infrastructure. Improving onboarding and providing real developer workflows would significantly reduce friction for builders and help the ecosystem grow faster.

Impact / Who Benefits

Developer

Pre-checks

• I've searched existing feature requests