Isolated cybercrime knowledge, built on Knolo.

Isolation at the organization and case level.

Knolo is the planned knowledge layer for CCRNet. We will build a separate organization pack for shared, authorized intelligence and separate case packs for case-specific reports, evidence, entities, and indicators.

Research privately first, then validate what can responsibly connect.

This is a sequence of research and prototyping priorities, not a release schedule. CCRNet has no expected launch date.

1. 01
   Build isolated Knolo packs

   Research independently scoped packs per organization and case so customers can work with their own authorized cybercrime data.

2. 02
   Develop deterministic cross-report search

   Prototype reproducible matches across selected case packs with source references and an auditable query scope.

3. 03
   Research human-reviewed connection suggestions

   Study how to propose possible relationships while requiring investigators to review supporting evidence and accept or reject each suggestion.

4. 04
   Evaluate opt-in sanitized intelligence sharing

   Research responsible ways for members to contribute selected, sanitized indicators and receive matches against intelligence shared by other members.

Search selected case packs deterministically.

This non-operational example helps guide current R&D. The host, authentication, schema, and endpoints are placeholders and may be substantially revised or abandoned based on research findings.

curl https://api.ccrnet.org/v1/search \
  -H "Authorization: Bearer $CCRNET_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "organization_id": "org_01",
    "case_ids": ["case_01", "case_02"],
    "query": "payment_identifier:wallet_123",
    "mode": "deterministic"
  }'

Training, development, and prediction with evidence in view.

CCRNet’s research focuses on cybercrime-specific data structures, dataset preparation, model training, evaluation, and responsible prediction. We are also studying whether Knolo-backed retrieval can keep model context appropriately scoped and traceable.

• Research retrieval restricted to organization and case packs the user is authorized to access.
• Preserve source references so investigators can verify search results, summaries, and predictions.
• Evaluate how review decisions can improve connection suggestions and model development.
• Treat model predictions as research outputs and investigative leads—not proof of crime or autonomous enforcement decisions.