tablevault 0.2.5

Centralized data repository for cross-process data filtering in Python.

TableVault

TableVault is a Python package for storing and querying workflow data with lineage tracking across scripts and notebooks.

It uses ArangoDB as the backend and gives you a single API (Vault) to:

• Store typed data lists (file, document, embedding, record)
• Track upstream/downstream dependencies between items
• Search by text, code provenance, and embedding similarity
• Coordinate long-running processes with safe pause/stop checkpoints

Documentation

You can find the full documentation at tablevault.org.

Installation

Install from PyPI:

pip install tablevault

Citation

If you use TableVault in research, cite:

• Zhao, J. and Krishnan, S. (2025). TableVault: Managing Dynamic Data Collections for LLM-Augmented Workflows. NOVAS @ SIGMOD.
  ArXiv: https://arxiv.org/abs/2506.18257

License

MIT License. See LICENSE.