LakeLogic
Loading…

Data contract platform

Trusted data products,
built by people and AI agents.

Design and deliver data products through executable contracts. Define ownership, enforce quality checks and monitor service levels.

Run a sample pipeline

Runs in Google Colab. Nothing to install.

Apache 2.0 open sourceRuns in your environmentDatabricksTechnology PartnerMicrosoftISV PartnerNVIDIAInception Program

Works with the stack you already run

  • Polars
  • DuckDB
  • Apache Spark
  • Databricks
  • Microsoft Fabric
  • Snowflake
  • BigQuery
  • Amazon S3
  • Azure ADLS Gen2
  • GitHub
  • GitLab
  • Azure DevOps
  • Bitbucket
  • Jira
  • Slack
  • Microsoft Teams
  • Power BI
  • Tableau

Why LakeLogic?

Go beyond tables and pipelines. Build data products with clear expectations and track whether they deliver.

Scan, profile, draft contracts

Discover assets and draft contracts that capture schemas, ownership and lineage.

  1. Assets and schemas discovered.
  2. Ownership and classification assessed.
  3. Dependencies declared, imported or inferred.
  4. Missing evidence listed.

Outcome

Dependencies are named up front, so impact analysis reads the lineage graph instead of a spreadsheet.

Understand your estate
marketing / google_analytics / silver / sessionsExample
Data productsmarketing domain
Data productOwnerClassificationDownstream
marketing.sessionsGrowth AnalyticsPII · restricted6 consumers
finance.revenue_dailyFinance OperationsConfidential11 consumers
crm.accountsSales SystemsPII · restricted4 consumers
iot.device_eventsPlatform EngineeringInternal2 consumers
Lineage resolved across 6 systems128 products
  • Defined contracts
  • Clear ownership
  • Quality controls
  • Service levels
  • Continuous monitoring

From data contract
to working pipeline.

Design, test and deliver data products on your platform.

  1. Design

    Draft contracts from your sources. Review schemas, quality rules and transformations.

  2. Test before delivery

    Run against synthetic data in Lakehouse Studio. Inspect accepted rows, failures and their reasons.

  3. Deliver

    Create a ready-to-run project for your platform, with contracts, pipeline code and jobs or workflows, opened as a pull request on your repository.

BatchMicro-batchStreaming

Batch, micro-batch and streaming workloads. Streaming runs on Apache Spark. Engine support

Explore a governed data mesh in action.

One set of data contracts builds domain-owned data products on Databricks, Microsoft Fabric and Snowflake. Open any project to see its pipelines, its quality checks and the rows they reject.

RideFlow architecture: YAML contracts and synthetic test data feed LakeLogic Core, which builds isolated domain products, shared multi-domain gold for consumers, quarantine for rejected rows and run metadata in your platform, with optional telemetry to the LakeLogic Platform.

  • Microsoft Fabric

    A governed data mesh on OneLake, with a lakehouse per domain. Fabric Data Pipelines run each domain from Bronze to Gold, with quality checks and quarantine from the contracts.

    Explore reference projectPlatform capabilities

  • Databricks

    A governed data mesh on Unity Catalog, with a catalog per domain. Databricks Workflows run each domain from Bronze to Gold, with quality checks and quarantine from the contracts.

    Explore reference projectPlatform capabilities

  • Snowflake

    A governed data mesh on Snowflake, with a database per domain. Snowflake Tasks run each domain from Bronze to Gold, with quality checks and quarantine from the contracts.

    Explore reference projectPlatform capabilities

RideFlow uses synthetic data; review each project’s prerequisites before deployment. LakeLogic Core executes supported contracts with Polars, DuckDB and Apache Spark. Compare engine support

Your production data runs in your environment.

Studio can also use samples you choose to upload for design and testing. Use synthetic or appropriately sanitised data.

  • Production runs in your environment

    LakeLogic Core processes production data in infrastructure you control and reports operational evidence to the platform.

  • You own the contracts

    Contracts remain version-controlled in your repository.

  • You approve changes

    Material changes pass through review and your delivery workflow.

  • Open execution

    LakeLogic Core is open source under Apache 2.0, so you can read the code that runs on your data.

Build with us.

Bring a real data engineering challenge. Work with the LakeLogic team as a design partner to shape, test and deliver a governed data product.

Explore the design partner programme

Get to know LakeLogic

What is LakeLogic?

A data contract platform. It discovers your data estate, helps you design and test data products in Lakehouse Studio, checks changes before merge, and monitors data quality and service levels.

What is a data contract?

A version-controlled file that holds a dataset’s schema, quality rules, owner and service levels. LakeLogic Core runs it, so the rules are enforced, not just documented.

Does my production data leave my environment?

No. LakeLogic Core processes production data in infrastructure you control and reports operational evidence to the platform. Studio can use samples you choose to upload for design and testing.

Which platforms does it work with?

Microsoft Fabric, Databricks and Snowflake each have a working reference project. LakeLogic Core runs contracts with Polars, DuckDB and Apache Spark.

Is LakeLogic open source?

LakeLogic Core, the engine that runs contracts, is open source under Apache 2.0. The platform adds discovery, Studio, review and monitoring.

How do I try it?

Run a sample pipeline in Google Colab with nothing to install, or book a demo and the team will set up an organisation with you.

Ready to build trusted data products?

See LakeLogic in action.