Kit 01
Founder operating memory
Capture decisions, summarize work, propose next actions, and keep a traceable memory of what changed.
A starter kit is not a consulting retainer. It is one reusable local workflow with a trigger, runner, evaluation, approval loop, artifact output, and audit trail.
Anatomy of every kit
Trigger
A schedule, file drop, or request starts the run.
Runner
A local agent does the work with the tools it is given.
Eval
Checks score the output before anyone relies on it.
Approval
A person reviews and signs off. Nothing ships itself.
Artifact
The run produces a real file, draft, or report.
Audit trail
Every run is logged and can be replayed later.
First starter kits
Kit 01
Capture decisions, summarize work, propose next actions, and keep a traceable memory of what changed.
Kit 02
Turn a source note into article, social posts, product copy, screenshots, and a final approval checklist.
Kit 03
Read sensitive files locally, draft an artifact, run checks, and preserve the audit trail.
Built on the NVIDIA stack
Orionfold is a member of the NVIDIA Inception program. Kits can run on customer-controlled NVIDIA GPU systems for private inference and evaluation, while preserving approval gates, logs, and artifacts for review.
See the DGX Spark field lab →© 2025 NVIDIA, the NVIDIA logo, NVIDIA DGX Spark, NVIDIA Inception, NVIDIA NeMo, NVIDIA NIM, and TensorRT-LLM are trademarks and/or registered trademarks of NVIDIA Corporation in the U.S. and other countries.