Skip to content
SCHEMAVORTEX

Self-managing data lakehouse

A governed data lakehouse,
in days.

SchemaVortex turns the data in your ERP, CRM and line-of-business systems into a governed lakehouse, with as much history as you choose to keep. It runs inside your own Azure subscription, on open Parquet and standard SQL. There are no pipelines to develop and no data-engineering team to hire.

Connects to

SQL Server PostgreSQL MySQL MariaDB Oracle SAP HANA IBM i Infor Data Lake REST & files

Built on

Azure Synapse Data Lake Storage Apache Parquet

Served to

Power BI Tableau Qlik Excel Python …any SQL client

The difference

A ready-to-run platform instead of a multi-year project.

Days, not years

Building a reliable lakehouse in-house takes one to three years and needs data engineers who are hard to hire and keep. SchemaVortex is delivered ready to run, and your existing BI team operates it.

Predictable cost

A product with costs you can plan around, instead of a build that keeps adding people. Microsoft bills the Azure it runs on to you directly, with no reseller markup and nothing idling between jobs.

Your data is not locked in

The only vendor-specific part is the platform software, and it runs in your tenant, operated by your team. Your data stays in open Parquet and is queried with standard SQL, so it stays yours with or without us.

The platform

The lakehouse, already built.

Everything between your operational systems and your dashboards, delivered as one governed product you run yourself.

SchemaVortex is deployed into the Azure subscription you already run. Your data is not sent to a provider's cloud.

Typical SaaS

Your data moves to the provider.

It is copied into the provider's cloud and held on their multi-tenant platform. You rent access to data that now lives somewhere you do not control.

SchemaVortex

The platform runs in your cloud.

It is deployed into your Azure subscription and run by your own team. The platform manages its own Azure resources, and there is no vendor access. Your data is governed where it lives and never leaves your tenant.

See how it runs in your tenant

Connect anything

Built-in connectors for the SQL databases and systems you already run, plus a developer SDK to push in anything custom.

  • SQL Server
  • PostgreSQL
  • MySQL
  • MariaDB
  • Oracle
  • SAP HANA
  • IBM i
  • Infor Data Lake
  • REST & files
  • …and many more
See everything it connects to

From source to dashboard

A complete lakehouse, delivered ready to run: ingestion, history, governance and serving in one product, with nothing to assemble or integrate.

Governed from day one

Four gates, data classifications that decide who reads each column, and a full audit trail, all active from the moment your first table is discovered.

See how governance works

Four eyes where you want them

Mart views, Vault tables and Sandbox views can change through proposals: checked against the current state, applied by a person entitled to apply them, under their own name, on the record.The AI Assistant proposes and never applies.

See how proposals work

AI

AI Chat explains your schema from metadata only, on your own Azure OpenAI. The AI Assistant lets a coding agent work for a user, under that user's own permissions, with every action logged.Neither sees more than the person asking.

How AI is governed

Run by your own team

Once the platform is set up and handed over, your BI team operates it and builds the models, all from the browser.

The data pipeline

Five steps from connection to query.

Set it up once, and SchemaVortex runs it from there, automatically.

See the storyboard
  1. Sources

    Connect

    Point SchemaVortex at any SQL source, and it discovers the tables, columns, types and keys automatically, with no connectors to build. Non-SQL sources come in through the Producer SDK.

    What it connects to
  2. Extraction

    Approve & classify

    Your data owners approve which tables and columns may enter the Vault, then classify the data only some people may read. Nothing lands unapproved.

    How approval works
  3. Vault

    Model

    Your BI team configures the Vault and the Mart. It decides which data keeps its history and shapes the masked, approved views your analysts work from.

    Explore the catalog
  4. Mart

    History

    From then on, every change is captured for you. The history is yours, and you can look back at any moment it still holds.

  5. Power BI / SQL

    Serve

    Ready-to-use data reaches the tools your team already uses, such as Power BI, Excel or any SQL client, with no servers to run in between. Access is decided as each query runs.

Highlights

One platform, end to end.

Governance

Governance built in, not bolted on.

Every column from every source passes four gates before anyone sees it. Access is decided at query time, so every query returns only what its reader may read, and there is no separate masked copy to keep in sync.

See how governance works
  1. AI GateColumns every AI session reads as empty, even where the person behind it may read them.
  2. Governance GateClassifications decide who reads the data.
  3. Approval GateNothing enters the Vault unapproved.
  4. Delivery GateA second set of classifications on Vault columns.

The Vault

Schema changes that do not break reports.

When a source adds a column or changes a type, one edit to the Vault table stores it as a new, versioned column and keeps the old, so the columns existing reports read do not change underneath them.

How the Vault works

Lineage

Every column's path is tracked.

Trace any column back to the source column it came from, or see everything a change would touch. The platform answers both from the views it runs, so view- and column-level lineage is always current.

Explore lineage

The Catalog

Browse, query and trace, in one catalog.

A single catalog across the Vault and the Mart, with a schema browser, a live SQL editor and one-click lineage on any column. It is built in, so there is nothing to deploy or scan.

Explore the catalog

Open by design

Your data. Your infrastructure.

Your data is open Parquet in your own storage, and your team queries it with standard SQL. The whole platform runs in your subscription, so there is nothing to migrate away from.

Why it is open
Data
Open Parquet, in your storage
Work product
Portable SQL
Platform
Run by your team, in your tenant

Data and work product stay yours. The platform software runs in your tenant, operated by your team.

Where we are

Based in Budapest, working across Europe.

Fizzcode Ltd., the publisher of SchemaVortex, is based in Budapest and works across Europe.