Skip to main content

What is Cogrion

Internal Documentation

This is the internal knowledge base for the Cogrion platform team


Cogrion is a managed data platform for data teams. It brings together SQL, pipelines, machine learning, and governance tooling in one place — provisioned and operated on the tenant's own cloud account.

The Cogrion Platform

Cogrion gives data teams the tools they need without the infrastructure burden. Instead of assembling and maintaining a data stack from scratch — selecting tools, configuring integrations, managing upgrades, and wiring authentication — tenants start from a working, opinionated setup and extend it from there.

The platform runs on a Bring Your Own Cloud model. Each tenant organization gets a dedicated environment provisioned into their own AWS or AliCloud account. Cogrion manages the control plane, the application catalog, and the web interface. Everything else — compute, storage, network, data — stays in the tenant's cloud. Data never transits through Cogrion infrastructure.

Built on Open Source

Cogrion is built on a curated set of open source technologies. Rather than building custom implementations, the platform packages, integrates, and manages well-established open source projects — and handles version updates, security patches, and configuration compatibility across the stack. Tenants get the ecosystem benefits of open source without the operational overhead.

Platform at a Glance

  • Workspace — Organize notebooks and files in a collaborative folder structure with sharing and scheduling
  • Catalog — Browse tables and schemas, assign ownership, trace lineage, and manage access
  • Workflow — Author and monitor DAG-based data pipelines with built-in scheduling and retry logic
  • SQL
    • SQL Lab — Write and run queries against data directly from the browser
    • Queries — Save, search, and re-run query history
    • Dashboard — Build interactive charts and share them across the team
  • Machine Learning
    • Experiments — Track training runs, compare metrics, and reproduce results
    • Features — Manage reusable feature groups with defined ownership and storage backends
    • Models — Register, version, and promote trained models through staging and production
    • Auto ML — Automated model selection and training with evaluation metrics
    • AI Gateway — Deploy models as REST endpoints — internal or external providers

See how different tenant teams use the platform on Common use cases


Find Your Path

You are...
Backend / API developerAuthentication · AI Gateway · Tenant isolation
Frontend / UI engineerSQL Lab · Dashboard · Auth flows
DevOps / InfraBYOC model · Architecture · Deployment lifecycle
Data engineerWorkflows · Catalog & lineage · Access policies
Data scientist / MLExperiments · Feature store · Model registry · AI Gateway
Sales / non-technicalUse cases · BYOC explained · Security summary