DEMO TENANTOperator-side data (staff, contacts, calendar, training, grants, complaints, SAR, four-eyes reviews) is illustrative — not a real Greentube record. Connector output, evidence-ledger anchors, RFC 3161 timestamps, Postgres RLS, regulator-source ingestion and LLM extractions are the real machine. Rows marked demo are seeded fixtures.
Georg Gubo
MGA · UKGC · ADM · RO-ONJN · SK-MFSR · DGOJ · GGL · SPI · NJ-DGE · MI-MGCB · PA-PGCB

Upcoming regulatory changes

LLM-extracted from regulator announcement pages across 12 jurisdictions. Review each impact before pushing tickets to your team.
Regulator:AllBR-SPACA-ON-AGCODE-GLUDK-SPMES-DGOJFR-ANJIT-ADMMGANL-KSARO-ONJNSE-SPIUKGC
Upcoming changes detected
0
Impacts pending review
0
LLM-processed
0

Changes

No upcoming changes detected in the current filter. The worker checks 12 regulator announcement pages daily; add KURAL_ANTHROPIC_API_KEY to enable LLM extraction. Manage the watch list at Regulator Sources.

How this works

KURAL watches official regulator announcement pages across all configured jurisdictions daily. When the page hash changes, the LLM extraction pipeline pulls the summary, effective date, and cited clauses out of the regulator's text via Claude Sonnet, then maps cited clauses to existing controls in your library via the source-citation index. Every extraction is logged with token cost + prompt hash in LlmCallLog.

AI-assessed fields show their confidence — anything below 85% waits in the per-impact review queue (open a change to triage). Once impacts are approved you can push them to Jira as future_compliance_gap findings — due date is set to the effective date minus 30 days.