aimarket-data-cap

Documentation

Data-as-capability. Private corpus → paid RAG-capability. Upload your private data corpus. It becomes a paid search capability. You earn 70% of every query. Doubles TAM — sell compute AND data. Snowflake-level monetization for AI marketplaces.

When to Use

• Law firm uploads 50k court decisions → legal.us-cases.search@v1, $0.05/query
• Medical publisher uploads research papers → medical.papers.search@v1, $0.10/query
• SaaS company uploads internal docs → company.kb.search@v1, $0.03/query
• Any data owner who wants to monetize their private corpus without giving it away

Installation

pip install aimarket-data-cap

Example

from aimarket_data_cap.data_capability import DataCapabilityRegistry

reg = DataCapabilityRegistry(platform_fee_pct=0.30)

# Law firm registers their corpus
dc = reg.register(
    owner_address="0xLawFirm",
    description="US court decisions 2020-2025, federal circuit",
    data_size_bytes=50_000_000, document_count=50000,
    query_price_usd=0.05,
    tags=["legal", "us-courts", "precedents"]
)

# Consumer queries — pays $0.05
result = reg.query(dc.capability_id, "precedent for breach of contract damages")
print(f"Price: ${result['price_usd']}")
print(f"Owner earns: ${result['revenue_split']['owner_usd']} (70%)")
print(f"Platform earns: ${result['revenue_split']['platform_usd']} (30%)")

# Data owner checks earnings
rev = reg.get_owner_revenue("0xLawFirm")
print(f"Total earned: ${rev['total_earned_usd']:.2f}")

Revenue Split

Default 70/30 — data owner gets 70%, platform gets 30%. Configurable per capability.

License

MIT · Maintained by AI-Factory