Querona use cases

Querona use cases#

Querona is plain Transact-SQL over your live data, so what you can build with it is open-ended — any job you can express as a query, a view or an endpoint over your sources is a Querona use case. The patterns below are some of the most common; treat them as a starting point, not a complete list.

Consolidate and modernize#

Bring scattered systems together behind one SQL Server–compatible surface, and change what sits underneath without disturbing the consumers on top.

  • Reuse what you already run — connect existing databases, files, SaaS and APIs as sources and query them as one, with no data movement. See Federation.

  • Integrate without building pipelines — model with views over live sources instead of copying data into a warehouse first; there is no ETL to build or maintain.

  • Keep legacy systems running — front an aging system with T-SQL, REST, GraphQL and MCP without changing the system itself, buying time to migrate on your own schedule.

  • Lift and shift to the cloud — applications keep one stable endpoint while you move the sources underneath, source by source, with nothing to re-point.

Serve and accelerate analytics#

Give reporting and analytics one governed model to read, and add speed only where it pays off.

  • A stable reporting backend — point Power BI, Tableau or Superset at one endpoint and one data model; because the physical engine underneath is decoupled and swappable, you can re-engine it without touching a report. Querona complements these tools rather than replacing them — it has no visualization of its own. See Querona for SQL Server users.

  • Accelerate selectively — materialize only the hot views, on the engine you choose, instead of warehousing everything, and redirect queries to pre-computed aggregates with query retargeting.

  • Self-service for analysts — pull any source into Python or a notebook through one endpoint, or join a large corporate dataset with a small local one straight from Excel. See Querona for analysts and self-service.

AI and the data lake#

  • Fresh data for AI agents — expose governed, always-current data to LLMs over the Model Context Protocol (MCP), without copying or mirroring it first: connect a source, declare the functions an agent may call, and answer questions from live data. See Querona for AI.

  • Compute over your data lake — read Parquet, Delta and CSV directly from the lake, process them in Querona or on an engine such as Apache Spark or StarRocks, and write the results back to the lake or return them to the client — no-copy, and not tied to any single engine.

Deliver fast, on your own terms#

  • Prototype and deliver fast — connect, model with views and serve in minutes: a dynamic data layer to start fast and iterate, and a good fit for data-hungry AI projects.

  • Run on your own infrastructure — on-premises, cloud or hybrid, on servers you control, with no SaaS lock-in and no data leaving your environment.

One manageable solution#

What ties these together is Querona’s core principle: express as much of a solution as you can in SQL — views and virtual databases over your sources — instead of as pipelines, connectors and scripts. Because the whole model is declarative SQL and metadata in one place, Querona takes on its upkeep: it tracks every dependency, keeps materialized copies refreshed, and lets you refactor, upgrade or move the model to a different engine as edits to SQL rather than a re-plumbing exercise — transparently to the consumers on top. You maintain one model, not a web of loosely-integrated moving parts, so there is far less to operate and keep in sync.

And because each pattern is just SQL, views and endpoints over your own data, the list above is open by design: if you can describe the data you want and where it should go, it is very likely something Querona can do.

Next steps#