Corint home

Build a shared
decision
semantic layer.

Connect risk data, policies, feature definitions, and decision history. Give analysts, AI agents, and decision engines a shared, traceable context.

DATA / Facts & signals

Connect your data.
Understand your business.

Make existing data assets discoverable, understandable, and usable—without creating a parallel inventory.

Data Facts & signals

Internal data

Operational databases · Warehouses · Transactions · Logs · Cases · Business outcomes

Third-party data

Credit bureaus · Identity · Devices · Lists · Open banking

01

Connect, without rebuilding

Reuse your existing data sources instead of maintaining another parallel inventory.

02

Scan progressively, guided by value

Start with structure, then samples, statistics, relationships, and semantics. Go deeper where value and scale justify it.

03

From investigation to runtime

Work reads the asset map for investigation and design. Decision resolves values at runtime using the same feature definitions.

Corint

Asset catalog / Core database

Transactions

Connected
Progressive scanningSemantics complete
  1. ✓Structure
  2. ✓Samples
  3. ✓Statistics
  4. ✓Relations
  5. ✓Semantics
transactions24 fields
Transactions fields.
Field / Business meaningTypeSample values
transaction_idTransaction IDstringtxn_demo_001
customer_idCustomer IDstringcust_demo_042
amountTransaction amountdecimal1,280.00
created_atTransaction timetimestamp2026-01-15 09:30

KNOWLEDGE / Definitions & explanations

Traceable knowledge.
Grounded understanding.

Preserve organizational knowledge and domain expertise. Reference them explicitly when needed—not as unbounded context.

Corint

Knowledge directory 03

Knowledge details v1.4

30-day transaction amount

Metric definitions

Sum successful transactions per customer over the past 30 days. Exclude reversals and refunds, using a consistent currency definition.

Source typeGit

risk-definitions / metrics / txn_amount_30d.yaml

Sync status✓ Synced
OwnerRisk strategy team

Knowledge Definitions & explanations

Internal knowledge

Metric definitions · Strategies · Rules · Models · Operating experience

Industry & domain knowledge

Industry methods · Regulatory policies · Risk patterns · Domain templates

01

Shared consensus. Private working materials.

Shared libraries hold organizational consensus. Private libraries keep individual working materials separate.

02

Every knowledge item has a source

Track the source—file, webpage, document, or Git—and sync status of each item, so definitions and explanations can be traced back.

03

Explicit references, not automatic context

Use knowledge through search or explicit task references.

THE CONNECTIVE LAYER

Decision Ontology

Bring facts, definitions, relationships, and knowledge into one computable business world.

01

Entities & relationships

CustomersAccountsApplicationsTransactionsProductsMarkets

Understand what exists in your business and how it connects.

02

Metrics & semantics

FieldsMetricsTagsDefinitionsLineageVersions

Give every metric a clear, consistent definition.

03

Decision assets

FeaturesRulesModelsStrategiesPoliciesRegulatory constraints

Connect business understanding to executable decision assets.

04

Decision memory

Operating experienceHistorical casesDecisions linked to business outcomes

Preserve the context behind decisions so experience stays traceable.

ONE DEFINITION. THREE CONSUMERS.

Define it once.
Use it across all three.

PeopleUnderstanding & governance

Analysts look up definitions in data assets and knowledge libraries, enrich semantics, and publish after review.

AgentInvestigation & design

Before writing SQL, investigating root causes, or generating features, Work reads the asset map and relevant knowledge—not unsupported assumptions.

Execution enginesRuntime execution

At runtime, Decision executes against the same feature definitions and strategy assets.

Same semanticsSame relationshipsSame version

Share understanding.
Respect every boundary.

Data connections

Connect to existing data sources and warehouses without duplicating the customer asset inventory.

Tenant isolation

Customer data, knowledge, and customer-specific ontologies are isolated by tenant.

Cross-customer reuse

Only industry methods and templates are reused across customers. No customer-specific assets are shared.

Knowledge governance

Agents cannot freely write or rewrite enterprise knowledge. Shared content requires clear ownership for maintenance and publication.

Give every decision a shared foundation.

Request a Demo