v1.0 Now in Beta

The Control Plane for
Data Products.

Define your data pipelines as declarative contracts. Enforce quality rules that catch impossible data before it ships. Govern the lifecycle from DRAFT to PUBLISHED.

Absolute Integrity. Governed Flow.

Get Early Access View Architecture

KAFKA KEDA POLARS ARROW

nyc-taxi.contract.json

{
  "apiVersion": "v1alpha",
  "kind": "DataProduct",
  "metadata": {
    "name": "nyc-taxi-yellow",
    "owner": "analytics-team",
    "status": "VALIDATED"
  },
  "spec": {
    // Quality rules
    "quality": {
      "behavior": "DROP_ROW",
      "rules": [
        {
          "check": "CUSTOM_SQL",
          "value": "dropoff > pickup"
        },
        {
          "check": "ALLOWED_VALUES",
          "column": "payment_type",
          "value": [1,2,3,4,5,6]
        }
      ]
    },
    // Aggregation
    "transformation": {
      "group_by": "VendorID",
      "agg": [
        {"SUM": "total_amount"},
        {"COUNT": "trips"}
      ]
    },
    // Delivery
    "delivery": {
      "target": "ADLS_PARQUET",
      "schedule": "*/10 * * * *"
    }
  }
}
                        

                                POST
                                /api/v1/data-products
                                201 Created
                            
→ Contract parsed. Version 2.1
→ Quality: 2 rules active
✓ Ready for production

Declarative

Define once in JSON. No imperative scripts.

Enforced

Bad rows get dropped, not shipped.

Auditable

Full SCD2 versioning. Every change tracked.

Capabilities

Everything you need for
production-grade data products.

From contract definition to delivery, Data Membrane handles the complexity so your team can focus on outcomes.

Contract-Based Pipelines

Define your entire data product as a declarative JSON contract. Sources, quality rules, transformations, and delivery—all in one versioned file.

Built-in Quality Gates

Row-level checks like CUSTOM_SQL catch impossible data. Batch-level rules act as circuit breakers.

Real-time Observability

Monitor every run, track quality violations, and get alerts. Full audit trail with SCD2 versioning—know exactly what changed.

Event-Driven State Machine

Enforce lifecycle transitions automatically. Only validated data moves from DEFINED → PUBLISHED.

Zero Idle Costs

KEDA-powered workers scale to zero when idle. Polars validates gigabytes in sub-second time. Pay only for what you process.

Universal Connectors

Ingest from ADLS, S3, Kafka, JDBC. Deliver to Snowflake, Delta Lake, Redshift, or Parquet. Swap components easily.

How It Works

From contract to production in minutes.

Define Your Contract

Declare sources, quality rules, transformations, and delivery targets in a single JSON file. Version it with your code.

Apply & Validate

Submit via API or UI. The state machine validates your contract and spins up workers automatically.

Ship with Confidence

Quality gates enforce your rules. Only valid data transitions to production. Full audit trail included.

The Control Plane for
Data Products.

Everything you need for
production-grade data products.

Contract-Based Pipelines

Built-in Quality Gates

Real-time Observability

Event-Driven State Machine

Zero Idle Costs

Universal Connectors

From contract to production in minutes.

Define Your Contract

Apply & Validate

Ship with Confidence

Connect everything.

Ready to govern your data mesh?

The Control Plane for Data Products.

Everything you need forproduction-grade data products.

Contract-Based Pipelines

Built-in Quality Gates

Real-time Observability

Event-Driven State Machine

Zero Idle Costs

Universal Connectors

From contract to production in minutes.

Define Your Contract

Apply & Validate

Ship with Confidence

Connect everything.

Ready to govern your data mesh?

The Control Plane for
Data Products.

Everything you need for
production-grade data products.