Solo Unicorn Club logoSolo Unicorn

Data Science Config

A CLAUDE.md for data science workflows with experiment tracking

$39Skill PackFor founders buying their first concrete result

What you can have running in the first 7 days

Get a ready-to-run system that replaces blank-page setup.
Ship a usable package with 7 included files and working structure.
Move from purchase to first setup in about 3 min.

What is Data Science Config?

CLAUDE.md for data science workflows. Jupyter conventions, experiment tracking, model evaluation protocols, data versioning, and reproducibility requirements.

Setup Time

3 min

Difficulty

Intermediate

Works With
claude-code

What's Included

  • CLAUDE.md
  • conventions/notebook-structure.md
  • conventions/experiment-tracking.md
  • conventions/data-versioning.md
  • templates/experiment-log.md
  • templates/model-card.md
  • README.md

Preview

CLAUDE.md
# CLAUDE.md — Data Science Config

## Notebook Conventions
- One notebook per experiment, named: YYYY-MM-DD-experiment-name.ipynb
- First cell: imports and config (no magic numbers in code cells)
- Last cell: summary of results and next steps
- Clear all outputs before committing

## Experiment Tracking
- Log every run: parameters, metrics, artifacts
- Use MLflow or Weights & Biases for tracking
- Never overwrite previous experiment results
- Tag experiments: exploratory, validation, production

## Reproducibility Requirements
- Pin all package versions in requirements.txt
- Set random seeds: numpy, torch, sklearn
- Document data source, version, and access date
- Include data preprocessing steps in pipeline (not notebook)

Installation Guide

Get up and running in under 5 minutes.

# Copy the config to your project root
cp data-science-config/CLAUDE.md ./CLAUDE.md

# Start a new session to load it
claude

Skill Pack. Pay once for the asset. Upgrade to implementation only when you want higher-touch help.

Share

Community acceleration

Bring your workflow into the Solo Unicorn community for sharper feedback, operator critique, and more visibility once the system is live.

Upgrade path

  • Start with this package and validate the workflow.
  • Add specialized skills or bundles once the core system is stable.
  • Use the community to sharpen positioning, demos, and feedback loops.

Need this adapted to your business?

Buy the asset first if you can run it yourself. If this workflow is business-critical or needs custom implementation, move into a sprint or fractional CIO advisory instead of guessing.

Discuss implementation →
Files included7
Setup time3 min
Difficultyintermediate

Tags

claude-mddata-sciencejupytermlexperimentsreproducibility