Branch Your Data, Query in Constant Time, Scale Forever

A Git-versioned SQL database built in C++17 with zero dependencies. Featuring the Aleph Filter — infinite-scaling probabilistic filtering based on VLDB 2024 research.

01

Git-like Versioning

Every write creates a commit. Branch your data for experiments. Merge changes back. Time-travel to any point. Your database, with version control built in.

02

C++17 SQL Engine

Modern C++ with zero external dependencies. Fast query parsing, efficient page-based storage, and a clean codebase. Built for learning and performance.

03

Aleph Filter

Skip disk reads with O(1) probabilistic filtering. Unlike Bloom filters, Aleph Filters expand infinitely without rebuilding. First non-Java implementation.

04

Copy-on-Write Storage

Branches share pages until a write forces a copy. Zero-cost snapshots, efficient branching, and time travel — all without duplicating data.

SQL meets version control

repono-cli
-- Create a table
CREATE TABLE users (
  id         INTEGER PRIMARY KEY,
  name       VARCHAR NOT NULL,
  created_at TIMESTAMP
);

-- Insert data
INSERT INTO users VALUES (1, 'Neel', 1703619600);

-- Commit changes
COMMIT 'Added first user';

-- View history
LOG;

-- Time travel
SELECT * FROM users AS OF 'abc123';

-- Branch
CHECKOUT -b feature;

Three clean layers

A layered architecture with clear boundaries. SQL at the top, version control in the middle, storage at the bottom. Each layer does one thing well.

01SQL Interface
02Version Control
03Storage Engine

Aleph Filter — To Infinity in Constant Time

The Aleph Filter is a probabilistic data structure based on VLDB 2024 research that enables O(1) membership queries and infinite expansion without rebuilding. Unlike Bloom or Cuckoo filters, it never hits a capacity wall.

RustFirst non-Java implementation
35k+Words in the research paper and tutorial
O(1)Query time, forever

Ready to branch your data?