Just ask.
Let RAW handle the rest

RAW introduces the NoDB approach to data processing: SQL-based queries that can access, transform and analyse even the most complex data sources, without any ingestion or preparation.

A query is all
it takes.

All actions are driven by the questions you ask, via advanced query optimisations performed just-in-time to transparently retrieve, cache, transform the data.

RAW builds the database from your questions, keeping what’s hot hot, and what’s cold cold.

high performance.

RAW employs sophisticated feedback-based optimisation loops to adaptively build and maintain high-performance code and large-scale caches.

The execution is optimised for the queries being performed, the source data being queried, and the output format being produced.

The data

Data is no longer just relational tables: logs for IoT, multi-dimensional arrays for AI, hierarchical data for Web Services.

RAW introduces a true polyglot data platform, capable of reading, joining and exporting the most complex structured data. And for when you need more, RAW can even mix other programming languages in a SQL statement.

build and share.

Built-in collaboration features allow users to build, augment and securely share datasets and data transformations.

So you get to do the work once only, find and use what your colleagues have built, and know exactly where each piece of data came from.

Security for the most demanding scenarios.

RAW is designed to operate in sensitive scenarios with stringent security constraints, such as the telecommunication and the financial sectors.

RAW runs on a Kubernetes cluster, where each query is isolated in a newly-generated Docker container, guaranteeing full resource isolation.

Engineered by an
award-winning team.

RAW Labs SA is co-founded by Anastasia Ailamaki, recepient of SIGMOD’s E. F. Codd Innovation Award for pioneering work on the architecture of database systems.

We are always looking for brilliant system engineers to build next-generation query engine architectures, performance optimisation and large-scale operations.

Join us!

RAW data platform

Data input

ORACLE, Microsoft SQL Server, MySQL, PostgreSQL, SQLite, Teradata, CSV, XML, JSON, HJSON, Excel, Text files, Log files, Parquet, Avro, HDFS, Web Services, Kafka, Arrow (in development)

Data output

JSON, HJSON, CSV, SQLite, Parquet, Avro, RAW-format, OpenAPI REST, Web Services

RQL: Extended SQL

The RAW Query Language (RQL) is an extended version of SQL with backward compatibility that allows you to query even the most complex hierarchical data without any prior preparation.


RAW supports multiple storage formats such as HDFS and S3.

Kerberos support

Out of the box support for Kerberos to securely access HDFS.


RAW is platform independent and can be deployed on OpenShift and AWS.

Native Python integration

RAW is fully integrated with Python allowing you to embed RAW queries in Python code and vice-versa.

Multiple client API support

RAW supports multiple client APIs such as Python, Java, Scala and Jupyter notebooks, plus our own easy to use command line client.

OpenAPI support

RAW fully supports OpenAPI providing you with language-agnostic interface for REST APIs into RAW and from RAW to external applications and microservices.

OAuth-based security model

Full integration with OAuth allows RAW to delegate secure access to its services.