Note
This document inventories the deliverable source tree for operators, integrators, and internal engineering.
Paladin is a proprietary trading intelligence application by Red Rook, LLC. It provides a PyQt5 workstation for charting, multi-timeframe market data, and machine-learning-assisted directional signals: a calibrated LightGBM ensemble, optional ONNX transformer inference, and engineered technical features over Yahoo Finance data via yfinance.
| File | Type | Role | Tech/Stats |
|---|---|---|---|
brain_model.pkl |
Default LightGBM stack from brains.py training |
||
paladin_brain.pkl |
Fallback brain if default is missing | ||
scaler.pkl |
Fitted RobustScaler |
Python 3.11 (see
requirements.txtcomments).RAM: 16 GB recommended for interactive ML workloads.
OS: Windows, macOS, or Linux.
pip install -r requirements.txtSome installs need a system TA-Lib library before the Python wheel; see upstream TA-Lib docs for your platform.
| Variable | Meaning | |
|---|---|---|
PALADIN_ORCA_PATH |
Absolute path to a GPT4All-compatible GGUF for humanize.py (overrides default cache). |
Baseline market data uses yfinance; no API key is required for standard Yahoo Finance access.
Warning
Paladin outputs analytics only. It is not investment advice, a solicitation, or a recommendation to buy or sell any security. Past performance does not guarantee future results. Operators must ensure compliance with applicable law and internal policy. Red Rook, LLC disclaims liability for trading losses.
© 2025 Red Rook, LLC. All rights reserved.
Document v1.0 · Last updated April 20, 2026
Icons: Font Awesome 6 Free (CC BY 4.0)