# How Cord Works

Cord turns operating data and team knowledge into defensible capital and resource allocation decisions through six stages.

## 1. Connect

Cord connects source systems, files, documents, and workflows. Common sources include CRM, ERP, finance, billing, project management, ticketing systems, roadmaps, codebases, spreadsheets, contracts, invoices, forecasts, and data-room materials.

## 2. Sanitize

AI agents clean, match, label, reconcile, and structure raw data. This creates high-fidelity operating context that can support analysis and reporting.

## 3. Ontologize

Cord models the business as a connected ontology. The ontology links systems, people, customers, contracts, products, projects, assets, costs, revenue, capacity, workflows, commitments, assumptions, and decisions.

## 4. Model

Teams ask allocation questions in natural language and run calculations on top of the ontology. Cord supports multi-hop graph traversals, aggregations, ROI models, payback calculations, predictive models, unit economics, capacity scenarios, and M&A diligence models.

## 5. Audit

Every material answer includes lineage. Users can trace the output back to source records, assumptions, transformation logic, and calculation paths.

## 6. Decide

Cord converts the analysis into decision-ready artifacts: sourced answers, business cases, board reports, dashboards, diligence packs, and recurring agent workflows.

## Why The Ontology Matters

The ontology gives humans and agents a shared semantic layer. It makes outputs drillable, scenarios reusable, and recommendations auditable. Instead of treating every analysis as a one-off spreadsheet, Cord compounds context across decisions.