Skip to main content

What Is RowOps

RowOps is a multi-tenant data ingestion platform. It processes tabular data through a configurable pipeline that runs client-side, with metadata and configuration stored server-side.

If you're integrating RowOps into your product, start with the External Developer Quickstart.

Core Architecture​

Processing Model​

Data processing occurs client-side. The pipeline executes in the browser (dashboard-assisted mode) or in Node.js (headless mode). See Data Handling for details on data flow.

Configuration Model​

Schemas, API keys, project settings, and import metadata are stored server-side. This separation allows:

  • Centralized schema management across clients
  • Tier-based feature enforcement
  • Audit logging of import events (metadata only)

Tenancy Model​

The system implements organization-based multi-tenancy:

  • Organizations contain projects
  • Projects contain schemas, keys, and import history
  • API keys are scoped to projects, not organizations
  • Role-based access control enforces permissions within organizations

What RowOps Does​

Ingests Tabular Data​

RowOps parses files (CSV, XLSX, and other supported formats) into tabular representations. Parsing occurs client-side.

Validates Against Schemas​

Schema-defined rules (type, required, regex, enumValues) are applied to each row. Validation produces explicit valid/invalid classifications with structured error information.

Applies Masking and Transforms​

Sensitive fields can be masked before export or delivery. Schema-driven transformations can reshape or derive column values.

Generates Column Profiles​

Statistical metadata (counts, distributions, min/max, inferred types) is computed for each column. Profiling supports data quality assessment before downstream delivery.

Delivers to External Systems​

The Sync module delivers validated data to user-configured endpoints. Delivery is client-driven with configurable batching and retry behavior.

Execution Modes​

RowOps provides two execution modes that share the same underlying engine:

Dashboard-Assisted Mode​

Users interact with a React-based UI to:

  • Configure schemas and mappings
  • Preview data and validation results
  • Trigger pipeline execution
  • Review import history

The dashboard does not execute data pipelines itself. Pipeline execution occurs in the browser.

Headless Mode​

Server-side or CLI applications invoke the pipeline programmatically via @rowops/headless:

  • No UI required
  • Secret key authentication
  • Configurable license modes (strict, demo)

Both modes apply the same schema definitions and tier-based limits.

Problems It Solves​

Client-Side Data Processing​

For applications that need to process data locally, RowOps runs the pipeline in the client environment. See Data Handling for details.

Schema-Driven Validation​

Instead of ad-hoc validation logic, schemas define validation rules declaratively. The same schema applies across browser and headless execution.

Structured Error Handling​

Invalid rows are classified with explicit error codes, row indices, and field references. Downstream consumers can inspect, export, or handle errors programmatically.

Tiered Feature Access​

Organizations can access features appropriate to their tier (Free/Pro/Scale/Enterprise). Tier enforcement occurs during license verification and at the WASM tier gate.

What This Page Does Not Cover​

  • Detailed pipeline stages (see Pipeline Overview)
  • Execution mode specifics (see Execution Models section)
  • Security and data handling claims (see Security section)
  • Module-level documentation (see Modules section)