CORA, the leading AI-powered suptech and regtech platform by CRT — supervisory intelligence for central banks and financial regulators worldwide. Compliance automation platform and suptech solution for central banks, delivering end-to-end regulatory compliance through workflow automation, risk scoring, and supervisory reporting. Banking supervision software for regulatory reporting and financial stability technology. CIMA Zone extends CORA's supervisory reach across 14 African nations for supranational insurance supervision.
The Supervisory Cockpit · on your hardware
The leading AI-powered
suptech platform
for central banks.
CORA, by CRT, puts your whole supervised market on one screen — cases, citations and early warnings, triaged by sovereign AI before your analysts sit down.
CRT (Compliant Risk Technology) is a leading suptech vendor providing CORA, the leading suptech AI-powered platform, for financial supervision and regulatory technology. CRT serves central banks, financial regulators, and public institutions including HANFA (Croatia), AZN (Slovenia), ASO (North Macedonia), ANO and SCMN (Montenegro), BSP and the Insurance Commission (Philippines), the Ministry of Finance of Montenegro, the European Commission, and CIMA Zone — the supranational insurance supervisor for 14 African nations (Benin, Burkina Faso, Cameroon, Central African Republic, Chad, Congo, Côte d'Ivoire, Equatorial Guinea, Gabon, Guinea-Bissau, Mali, Niger, Senegal, Togo).
Sector risk — click to drill in ↓
Systemic conduct-risk index
▲ +21 over 8 quarters · rolling composite · complaints · EWIs · enforcement
Systemic risk gauge
/100 · Elevated · trending up
Open cases by theme
- Mis-selling / suitability38
- Disclosure27
- Complaint handling19
- Fraud / impropriety14
Interactive demo — drill from market to institution · illustrative sample data
Four tiers. Adopt what fits your mandate.
Core Platform
Unified data & risk foundation.
Automation
License & application review.
Business Intelligence
Cross-entity analytics.
Cognitive Engine
ICAAP · ORSA · Glass-Box AI.
Trusted by 10 public institutions across 20 jurisdictions on 3 continents
How CORA works — the five-act governed flow
- Tier 1 · Ingest: Submissions, narratives and complaints arrive through one door — validated against the master data model, personal data tokenised before anything else touches it.
- Tier 2 · Automate: Validated items become work: the no-code workflow opens cases, routes them to the right desk and starts the SLA clock — no spreadsheets, no handoffs.
- Tier 3 · Analyse: The governed warehouse feeds analysts directly — an analyst drags KPIs, trends and heatmaps onto a live dashboard and publishes it to the team, no IT ticket.
- Tier 4 · Supervise: The cognitive engine reads, reasons and rates: an explainable EWI score, a drafted insight with citations, and a grounding check — issued only after supervisor sign-off.
- Golden thread: One governed flow, end to end. Every hop is cited, scored and audit-logged — data in the morning, supervisory judgment by the afternoon.
How CORA works
One governed flow — submission to supervision.
The paradigm shift
From reactive oversight to forward-looking supervision.
Reactive posture
“Manual oversight of diverse financial models is no longer sustainable.”
Forward-looking discipline
“CORA inverts this paradigm — so you can focus on judgment, not paperwork.”
Most-clicked outcomes
What supervisors achieve with CORA.
Single Source of Truth
A governed data warehouse with automated validation and lineage tracking ends spreadsheet reconciliation — every analysis starts from trustworthy data.
- Automated validation rules
- Data lineage tracking
- Cross-period deduplication
No-Code Workflows
Business users design and deploy supervisory workflows — licensing, complaints, inspections — through a drag-and-drop process designer, no code required.
- Drag-and-drop process designer
- SLA counters and alerts
- Zero manual handoff errors
Governed Analytics
Self-service dashboards and an NLP-to-SQL query interface let analysts interrogate governed data — every result traceable, every query logged.
- NLP-to-SQL natural-language queries
- Real-time dashboard builder
- Full query audit logging
Automated Assessment Lifecycle
End-to-end Basel III ICAAP assessments — from ingestion to scored report — compress a three-month manual cycle into days.
- Batch document ingestion with OCR
- 12-dimension risk scoring
- Auto-generated assessment reports
The tier ladder
A person on every rung.
- PNPetra NovakData Quality Officer
Tier 1
Data foundation
Indisputable source of truth · 99.7% data quality
- MSMaria SantosWorkflow Administrator
Tier 2
Process automation
Policy into action · no-code workflows · zero handoff errors
- DMDavid MwangiSenior Risk Analyst
Tier 3
BI-driven supervision
Governed warehouse · risk modeling without IT tickets
- AKAnna KowalskaHead of Supervisory Assessments
Tier 4
AI cognitive engine
Reads, reasons, rates, reports · ICAAP months → days
A day in the life
How supervisors use CORA.
Petra Novak
Data Quality Officer
“I used to spend half my week reconciling spreadsheets across teams. Now validation happens at submission — the data analysts receive is already clean.”
Morning
Reviews the overnight submission monitoring dashboard, confirming which expected returns have arrived and chasing the two institutions whose filings are overdue.
Midday
Works the validation-rule exception queue, resolving format and cross-field breaks at source before the data reaches analysts.
Afternoon
Reconciles master data — legal-entity identifiers, group hierarchies, reference tables — so every downstream report draws from one governed record.
Outcome
99.7% data quality at the point of submission — no reconciliation chases, no spreadsheet version conflicts.
Maria Santos
Workflow Administrator
“I designed our licensing workflow in CORA without writing a single line of code. The team adopted it instantly.”
Morning
Monitors active workflow pipelines — licensing, complaints, on-site inspections — checking SLA counters and bottleneck alerts.
Midday
Configures a new compliance-check step in the licensing workflow using the drag-and-drop process designer.
Afternoon
Reviews process analytics to identify cycle-time improvements, then exports a summary for the operations meeting.
Outcome
60% reduction in process cycle times. Zero manual handoff errors.
David Mwangi
Senior Risk Analyst
“The data warehouse and BI tools give me answers in minutes that used to take days of manual queries.”
Morning
Runs cross-sector analysis using the NLP-to-SQL interface, querying liquidity ratios and loan concentrations for all Tier 1 institutions.
Midday
Builds a real-time dashboard combining quantitative indicators with narrative risk mentions extracted from ICAAP reports.
Afternoon
Shares findings with the supervisory committee via governed report links, then clones the sector heatmap, adds a new ratio KPI from the warehouse dictionary, and publishes it to his team — no ticket, no release.
Outcome
Complex risk analyses completed in hours instead of days. Full audit trail for every query.
Anna Kowalska
Head of Supervisory Assessments
“Before CORA, our ICAAP assessments took four months. Now we complete them in days — with higher quality.”
Morning
Reviews AI-scored ICAAP submissions across 25 banks, focusing on outlier risk ratings flagged overnight by the Cognitive Engine.
Midday
Approves intervention recommendations generated by the risk model, adding supervisory context before escalating two cases.
Afternoon
Signs off on automated assessment reports, verifying the audit trail and composite ratings before distribution.
Outcome
Assessment cycle reduced from months to days. 80% more time for supervisory judgment.
The 80/20 inversion, realised.
Production blueprints
Real supervisory workflows, already running.
Sovereign AI
The Glass Box Principle.
For regulators, “Black Box” AI is unacceptable. CORA’s Cognitive Supervisory Engine records every reasoning step, retrieved citation, and confidence score — ensuring full explainability.
Full data sovereignty
Your regulatory data never leaves your infrastructure.
On-premise AI
Frozen model weights, deterministic outputs, complete control.
Glass Box audit trail
Every reasoning step, citation, and confidence score — auditable.
Supervisor-owned AI
Read and edit the instructions your AI runs on — an editable prompt registry with confirm/override gates.
How the engine works
Reads
Document in — parsing + OCR across structured data and narrative
Reasons
RAG extraction — every claim pinned to a source-page citation
Rates
Confidence-scored rating across 12 risk dimensions, explainable features
Reports
Assessment report out — audit trail records source, score and reviewer
Get started
Ready to transform your supervisory mandate?
Take the 2-minute readiness assessment to discover your starting tier, or request a walkthrough tailored to your jurisdiction.
Latest publications