byzer-org/byzer-lang

Health v2 · maintenance only

Adoption not part of the score

1835 stars · 537 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Byzer (formerly MLSQL) is a low-code, SQL-like distributed programming language and engine for data pipelines, analytics, and AI, built around the principle that everything is a table. It runs on the JVM (originally on Spark) and ships with built-in algorithms and extensions so users can build data apps without directly interacting with the underlying compute framework.

Use cases

• build etl data pipelines with a sql-like language
• run distributed data analytics on big data
• train machine learning models using sql-like syntax
• unify data engineering and ai workflows in one language
• process external apis as tables for analysis
• build notebook-based data products on spark

When to choose

• your team knows sql and wants low-code data pipeline and ml tooling
• you want a unified language for etl, analytics, and ai on spark-scale data
• you prefer everything-as-a-table semantics over stitching multiple tools

When to avoid

• you need a general-purpose programming language with a large ecosystem
• your workloads are small and a pandas/dbt-style stack would suffice
• you need actively rapid development - the project's latest release is from mid-2024

Facets

framework · maturity maintenance

programming-language interpreter etl machine-learning data-science big-data machine-learning data-science programming-languages jvm cloud self-hosted sql-like-dsl low-code mlsql distributed-computing spark data-pipeline analytics sql data-engineering docker linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="byzer-org/byzer-lang")