> ## Documentation Index
> Fetch the complete documentation index at: https://develop.cotality.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Enterprise data teams

> How Cotality's enterprise data teams govern, engineer, and deliver the trusted data assets that underpin every API, AI model, and data product on this portal.

Within Cotality, enterprise data teams act as the centralized engine that turns raw, disparate data into trusted, secure, and compliant consumption-ready assets. For developers and integration teams, understanding this group explains where the data comes from, how it is governed, and what rules apply when building on top of it.

***

## Who enterprise data teams are

Enterprise data teams comprise four functional groups: Enterprise Data Management (EDM), Data Analytics, Data Science, and Data Platform. Together they operate as the "hub" in a hub-and-spoke stewardship model that spans every business domain.

<CardGroup cols={2}>
  <Card title="The hub — centralized data teams" icon="building">
    Enterprise Data Management (EDM), the Chief Data Office (CDO), and Data Platforms. They provide standardized templates, data-engineering pipelines, central tooling (including the Collibra Data Catalog), and compliance audits.
  </Card>

  <Card title="The spokes — federated domain teams" icon="diagram-project">
    Domain-embedded data owners, stewards, and subject matter experts (SMEs) across business units — Real Estate, Data Solutions, Mortgage, Insurance — who manage specific functional data assets.
  </Card>
</CardGroup>

***

## What enterprise data teams do

Enterprise data teams manage data across its entire lifecycle — from initial sourcing through productization — organized around four pillars.

### 1. Data stewardship and knowledge engineering

* **Metamodel curation** — Defining standardized business vocabularies, enterprise ontologies, and technical schemas
* **Critical data element (CDE) tracking** — Identifying and tagging key business elements (KBEs) vital for business operations, regulatory reporting, and downstream product integrations
* **Data lineage mapping** — Constructing end-to-end lineage maps that trace data from its point of origination through all transformations to final consumer endpoints

### 2. Total quality management

Continuous profiling and automated validation across four core dimensions:

| Dimension        | What it measures                                                 |
| ---------------- | ---------------------------------------------------------------- |
| **Accuracy**     | Alignment to verified target sources and realities               |
| **Completeness** | All required attributes and value sets are present               |
| **Timeliness**   | Data is delivered as rapidly as downstream processes require     |
| **Consistency**  | No conflicting facts exist across target and source environments |

Initial and periodic profiling detects structural anomalies, null sets, and formatting errors. Root-cause analysis and remediation pipelines prevent bad data from propagating downstream.

### 3. Data protection, privacy, and access

* **Data classification** — All incoming data assets are categorized into security levels: Public, Confidential, and Non-Public Personal Information (NPI)
* **Privacy-by-design guardrails** — Row-level and column-level masking, automated Personal Information (PI) and Sensitive Personal Information (SPI) filtering, and exfiltration defenses embedded into AI/ML data flows
* **RBAC and least privilege** — Role-based access controls ensure developers and platforms interact only with the minimal dataset required for their functions

### 4. Integration, processing, and fulfillment

Raw external data is cleansed and standardized, then engineered into three classes of advanced derivative assets:

| Asset class  | Description                                                                                                               |
| ------------ | ------------------------------------------------------------------------------------------------------------------------- |
| **Blended**  | Combines multiple source datasets                                                                                         |
| **Modeled**  | Houses pure algorithmic outputs — predictions, scores, classifications                                                    |
| **Enriched** | Merges modeled outputs with source or blended data to build high-value, multi-dimensional structures (e.g., Property 360) |

***

## Data domains under governance

Enterprise data teams govern 13 core data domains organized into three logical groups.

### Master domains

| Domain          | Managed data assets                                                                     |
| --------------- | --------------------------------------------------------------------------------------- |
| **Individual**  | Customer identity data, retail investor profiles, and global opt-in/opt-out preferences |
| **Institution** | Institutional identities, partner IDs, and corporate investor details                   |
| **Product**     | Normalized hierarchy of Cotality products, packages, and crosswalk indexes              |

### Transactional domains

| Domain                    | Managed data assets                                                                                       |
| ------------------------- | --------------------------------------------------------------------------------------------------------- |
| **Credit / underwriting** | Credit risk data, bureau reports, loss forecasting metrics, and model outputs spanning the loan lifecycle |
| **Funding / platform**    | Investor funding details, platform balances, and fulfillment requirements                                 |
| **Servicing**             | Current loans, repayment histories, delinquent loan metrics, and collection activities                    |
| **Customer contact**      | Customer interactions — telephony, chat transcripts, complaints — and product analytics                   |
| **Marketing**             | Lead-generation logs, digital campaign clickstream data, and outreach targets                             |

### Corporate domains

| Domain              | Managed data assets                                                           |
| ------------------- | ----------------------------------------------------------------------------- |
| **Finance**         | General ledger data, accounting positions, and financial regulatory reporting |
| **Enterprise risk** | Risk assessments, key risk indicators (KRIs), and internal audit findings     |
| **Technology**      | Infrastructure maps, tech-asset inventory, and operational systems            |
| **HR and legal**    | Workday employee metrics, legal matters, and attorney-client privilege data   |

***

## Guidelines for developers integrating data

<Warning>
  **Permissible use certification** — No consumption-ready data asset is eligible for API integration or product deployment until the designated data owner formally certifies that it complies with all legal, regulatory, and contractual sharing restrictions.
</Warning>

<Note>
  **AI input validation** — If you are building AI systems or utilizing MCP servers, all data pipeline training sets and configuration files (e.g., YAML) must be formally logged in the Metadata Catalog.
</Note>

<Note>
  **Derivative data rules** — Any predictions, classifications, or synthetic attributes generated by downstream machine learning models are classified as new data assets and must immediately be registered under Cotality's data governance policy.
</Note>