featcat

CI PyPI Python License

AI-Powered Feature Catalog for Data Science Teams

Tiếng Việt

featcat is a lightweight Feature Catalog designed for Data Science teams. It is not a Feature Store (no online serving) — it's a metadata management tool with an AI layer for searching, documenting, and monitoring feature quality.

The Problem

  • Features scattered everywhere: Parquet files stored across local disks, S3, and MinIO — nobody knows what features exist
  • Missing documentation: Dataset columns have no descriptions; new team members don't know what avg_session_duration means
  • Hard to find the right features: Starting a new project (e.g. churn prediction) with no idea which features are already available
  • Undetected data drift: Feature distributions change silently until model performance degrades

Key Features

Module Description Phase
Catalog Register data sources, scan Parquet to auto-extract schema + stats 1
AI Discovery Describe a use case → AI recommends relevant features + suggests new ones 2
Auto-doc LLM automatically generates documentation for each feature 2
NL Query Ask in natural language (English or Vietnamese), AI finds relevant features 2
Monitoring PSI drift detection, null spikes, range violations 3
TUI Terminal UI with dashboard, feature browser, AI chat 3
S3 Support Read Parquet directly from S3/MinIO — never copies data locally 1
Caching Cache LLM responses to speed up doc generation and NL queries 3

Quick Start

# 1. Clone and install
git clone https://github.com/codepawl/featcat.git && cd featcat
uv venv && source .venv/bin/activate
uv pip install -e ".[dev]"

# 2. Initialize catalog
featcat init

# 3. Register and scan a data source
featcat source add device_perf /data/features/device_performance.parquet
featcat source scan device_perf

# 4. Browse features
featcat feature list
featcat feature info device_perf.cpu_usage

# 5. (Optional) Enable AI features — requires Ollama
ollama serve &
ollama pull lfm2.5-thinking
featcat discover "churn prediction for telecom customers"
featcat ask "features related to user behavior"

TUI (Terminal UI)

uv pip install -e ".[tui]"
featcat ui

Keybindings: D Dashboard | F Features | M Monitor | C Chat | Q Quit | ? Help

System Health Check

featcat doctor

[x] Python 3.10+
[x] SQLite catalog exists (catalog.db)
[x] Ollama running at localhost:11434
[x] Model lfm2.5-thinking available
[x] 14 features registered
[x] 10 features have docs (71.4%)
[ ] 2 features have drift warnings

Tech Stack

  • Python 3.10+ | SQLite (metadata only, never copies data)
  • Typer + Rich (CLI) | Textual (TUI)
  • PyArrow (Parquet schema + stats) | s3fs (S3/MinIO)
  • Ollama (local LLM) | Pydantic (models + config)

Project Structure

featcat/
├── catalog/        # Models, DB, scanner, storage backends
├── llm/            # LLM abstraction (Ollama, llama.cpp)
├── plugins/        # Discovery, Autodoc, Monitoring, NL Query
├── utils/          # Prompts, catalog context, statistics, cache
├── tui/            # Textual TUI (screens, widgets)
├── config.py       # Pydantic settings
└── cli.py          # Typer CLI entry point

License

MIT

Project links

Data verified by PyPI on Apr 6, 2026
Data provided by the project maintainers, verified at the time the release was uploaded to PyPI.

Key dates

PyPI data
Data sourced directly from PyPI's database.
  • Released:
Latest release

1 maintainer

PyPI data
Data sourced directly from PyPI's database.
Avatar for nxank4 from gravatar.com nxank4

Credits

Author: Codepawl

GitHub Statistics

Data verified by PyPI on Apr 6, 2026
The GitHub source repository was provided by the project maintainers and verified by PyPI at the time of upload. Stars, forks, and open issues/PRs are derived from that repository and have not been independently verified.

License

MIT License (MIT)

Requires

Python >=3.10

Provides Extra

all dev s3 server tui

Tags

ai data-science feature-catalog feature-store mlops

Classifiers

Report project as malware

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

featcat-0.2.4.tar.gz (346.4 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

featcat-0.2.4-py3-none-any.whl (101.4 kB view details)

Uploaded Python 3

File details

Details for the file featcat-0.2.4.tar.gz.

File metadata

  • Download URL: featcat-0.2.4.tar.gz
  • Upload date:
  • Size: 346.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.7

File hashes

Hashes for featcat-0.2.4.tar.gz
Algorithm Hash digest
SHA256 343b17640ca74399793f11f6e4b92f364b1d8a076f1a9af9a242dedf7f534805
MD5 739ed69dcc0703985717501fc28ebcbc
BLAKE2b-256 d9a100986f4689941d600f50c13029027e40f41b43d9ee37c35c2d89b3715d0f

See more details on using hashes here.

Provenance

The following attestation bundles were made for featcat-0.2.4.tar.gz:

Publisher: publish.yml on codepawl/featcat

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file featcat-0.2.4-py3-none-any.whl.

File metadata

  • Download URL: featcat-0.2.4-py3-none-any.whl
  • Upload date:
  • Size: 101.4 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.7

File hashes

Hashes for featcat-0.2.4-py3-none-any.whl
Algorithm Hash digest
SHA256 0d27463a005226e80ea315a7d6b3290ccbd1c8b545f3af739469483d38db0cd6
MD5 8ed07401cdd754a14c7a837105ef77d5
BLAKE2b-256 2e7c9cbe5b3de178f7af0edfb1215c5edb1af19d9d32f9ee15721b7c27091802

See more details on using hashes here.

Provenance

The following attestation bundles were made for featcat-0.2.4-py3-none-any.whl:

Publisher: publish.yml on codepawl/featcat

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Release history Release notifications | RSS feed

This release

0.2.4 This release

2 files

0.2.3

2 files

0.1.0

2 files

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page