Skip to content
Projects

data-infra · open-source · python

Data Curator

Open-source (MIT) component library behind Kaxanuk's data stack: downloads, validates, homogenizes, and combines multi-provider equity and fundamental data with US GAAP-based tags and point-in-time discipline.

https://github.com/alanvaa06/Data-Curator

Overview

Data Curator is the open-source component library that powers Kaxanuk's data infrastructure — a standalone development by Alan, not the company itself. It downloads, validates, homogenizes, and combines equity and fundamental data across multiple providers, exposing them through a US GAAP-based tag taxonomy with point-in-time discipline. Released under MIT, it runs as a Python library, a Docker image, or through a browser-based parameter panel.

What it does

  • Provider-homogenized tag names on a US GAAP-based taxonomy — switch data providers without changing code.
  • Automatic validation of market and fundamental data — discards impossible datasets (e.g., price high below low) and non-point-in-time amendments.
  • User-defined calculated feature functions without requiring NumPy or Pandas.
  • Output to CSV, Parquet, or in-memory Pandas DataFrames.
  • Extensible architecture: plug in custom data providers, feature combinations, and output handlers behind stable interfaces.
  • Browser-based parameter panel for configuration, plus a Docker image for containerized runs.

Stack

Python 3.12-3.14 · pandas · Docker

Status

Open source, MIT license.

GitHub →