Frequently asked questions
These answers are designed to help you evaluate fit quickly. If your use case is unusual or you have formal governance requirements,
HFA can scope deliverables to align with your workflow.
Licensing & usage
How licensing works
Do you sell individual clips or consumer licenses?
No. HFA is designed for research and product development teams. Licenses are scoped around datasets (or dataset sets),
with terms aligned to your intended use case and deployment environment.
What’s the difference between Research, Startup, and Enterprise tiers?
Tiers reflect differences in usage scope, documentation depth, and governance support—not differences in audio quality.
Foundations audio is included at every tier, and the Proteus Standard™ integrity framework applies across deliveries.
How are licenses scoped?
Licenses are typically scoped by (1) use case (research, product R&D, production deployment), (2) dataset coverage,
(3) term and scope (internal use, distribution, renewal), and (4) any required compliance or audit alignment.
Can we license multiple datasets as a bundle?
Yes. Many teams start with a focused evaluation set, then expand into a multi-dataset agreement as coverage needs become clearer.
Bundling is common for long-term partnerships.
Deliverables
What you receive
What is included in every HFA dataset?
Every dataset includes Foundations: rights-cleared recordings, consistent organization, and structured metadata that supports
filtering, selection, and repeatable ingestion workflows.
What’s the difference between a Preview and a Full Dataset?
Previews are designed for evaluation—representative audio and metadata examples to validate fit. Full datasets include the complete package
and the full Proteus Standard™ artifact set appropriate to the delivery and tier.
What is the Orpheus Suite?
Orpheus is an enterprise enrichment layer built on top of Foundations. It adds modeling-optimized structure—such as instruction-style
examples, alignment-friendly metadata fields, and optional exports for training pipelines.
What file formats and metadata formats do you provide?
Datasets are delivered with clean, consistent audio exports and structured metadata designed for programmatic ingestion. When needed,
HFA can provide additional exports (e.g., JSONL/Parquet) for training and evaluation workflows.
Rights & provenance
Defensibility and audit readiness
How are rights handled?
HFA datasets are recorded and delivered with a compliance-first approach. Performers are rights-cleared, provenance is documented,
and datasets are structured for long-term auditability.
What is the Proteus Standard™?
The Proteus Standard™ is HFA’s integrity and provenance framework. It links sessions to deliverables, provides verification artifacts,
and supports defensibility in governance, compliance review, and diligence.
Can we use HFA datasets in production systems?
Yes—production deployment is supported under appropriately scoped licensing. If your organization has formal review requirements,
HFA can align documentation and deliverables to the level of rigor you need.
Practical workflow
Getting started smoothly
How do we choose the right dataset(s) to start with?
Start with your target behaviors and the smallest coverage set that can validate them. If you’re unsure, HFA can recommend a focused
evaluation set based on your model type and goals.
Do you provide guidance on ingestion and evaluation?
Yes. Teams typically begin with a brief licensing conversation, then align on dataset selection, metadata depth, and evaluation slices.
Enterprise workflows can include additional coordination for governance or compliance needs.
Can we request a specific dataset or recording plan?
In some cases, yes. If you need a specific instrument, articulation set, or coverage profile, HFA can discuss feasibility and
timeline as part of a partnership conversation.
