Claude Code · Data Science

understand first.
code second.

A Claude Code setup that turns your IDE into a Data Science dojo. 12 steps. One at a time. The code is not the lesson — the thinking is.

claude-code — loan-eligibility/
/start Scanning project directory... train.csv — 614 rows × 13 cols PROBLEM.md found Reading problem statement... Business: Loan eligibility prediction Target: Loan_Status (binary) Metric: Accuracy ────────────────────────────── Before we look at any code — What type of ML problem is this? Classification, regression, or something else? Take a guess. 🤔

The Socratic Ladder

Claude never jumps to the answer. It climbs — asking, hinting, scaffolding, and only revealing as a last resort. The struggle is the lesson.

Level 1
💬
ASK
Question first. Always.

"What do you think we should do here? Why?" — Claude never opens with code. It opens with curiosity.

Level 2
💡
NUDGE
Hint. No code yet.

A conceptual pointer toward the answer. Analogies, mental models, things to think about — but still no code.

Level 3
🧩
SKELETON
Fill the blanks.

Partial code with gaps for you to complete. The structure is there. You supply the thinking.

Level 4
REVEAL
Last resort. Every line explained.

Full code with a line-by-line breakdown. Only reached after genuinely working through every other level.

Never skip to Reveal. Never. If you can't explain it, you haven't learned it yet.

Everything a self-taught DS student needs.

Not an autocomplete tool. Not a shortcut. A learning environment built to make you actually understand what you're doing — not just get outputs.

🗺️
12-Step Roadmap

Every dataset follows the same path: Understand → Explore → Prepare → Model → Improve. The order never changes.

🧠
Socratic Teaching

Asks before it tells. Hints before it solves. You never run code you can't explain back.

💾
Session Persistence

Auto-saves to PROGRESS.md after every step. Close your laptop, come back tomorrow, pick up exactly where you left off.

📄
Problem Context

Drop a PROBLEM.md with business context and Sensei teaches within that frame — data dictionary, metric, everything.

🔍
Auto Domain Detection

Reads your data, detects fintech / health / gaming / ecommerce, and adapts every example and analogy accordingly.

🪜
The Socratic Ladder

Stuck? Ask → nudge → skeleton → reveal. Never jumps straight to the answer — makes you work for it.

📓
Cell Discipline

Every notebook cell needs a WHAT and WHY comment. Builds intentional habits instead of running cells blindly.

Gate Checks

Can't move to the next step until you prove you understood the current one. Learning over output, every time.

12 Steps. One at a Time.

The order never changes. Steps shrink or expand based on your data — but the sequence is fixed, because the thinking process behind DS is universal.

Phase 1 UNDERSTAND
  • 01
    Dataset Orientation
    What is this data? What's the business question?
  • 02
    Problem Identification
    Classification? Regression? What's the target?
Phase 2 EXPLORE
  • 03
    Data Profiling
    Structure (shape, dtypes, columns) + quality (nulls, duplicates, issues)
  • 04
    EDA
    Distributions, relationships, outliers
Phase 3 PREPARE
  • 05
    Train-Test Split
    Always before any transformation
  • 06
    Feature Engineering
    New features, transformations, domain signals
  • 07
    Preprocessing Pipeline
    Imputation, encoding, scaling — fit on train only
Phase 4 MODEL
  • 08
    Baseline Model
    Simplest model first. Establish a benchmark.
  • 09
    Model Selection
    Try 2–3 models. Compare with cross-validation.
  • 10
    Evaluation
    Right metrics for the right problem.
Phase 5 IMPROVE
  • 11
    Hyperparameter Tuning
    Grid or random search. Understand what changes.
  • 12
    Final Insights
    Feature importance, business interpretation.

6 Commands. Everything You Need.

Type these in Claude Code. That's it — no configuration, no plugins. Just open your project and start learning.

/sensei
Initialize a Classroom

Run this once in any folder to turn it into a Sensei classroom. Writes CLAUDE.md to the folder for future sessions and primes the current session immediately. Re-run it anytime — older classrooms auto-upgrade to the latest ruleset. Then run /start.

bootstrap opt-in auto-upgrade
/start
Begin a Session

The entry point for every session. Reads PROBLEM.md, scans for data, shows your 12-step roadmap, and resumes right where you left off if PROGRESS.md exists — including flagging topics that tripped you up last time so they get revisited.

session start resume roadmap weak spots
/next
Move to the Next Step

Gate-checks your understanding with one targeted question. If you answer well, moves forward; missed answers get logged as weak spots to revisit later. Auto-saves PROGRESS.md silently every time — and after Step 12, runs a graduation retrospective.

gate check auto-save graduation
/checkpoint
Test My Understanding

3 targeted questions for your current step — one at a time, matched to your problem type. One question aims at a logged weak spot; nail it and the weak spot is cleared as mastered. Own the concept, not just the output.

quiz comprehension clears weak spots
/roadmap
Show My Progress

Visual progress tracker showing every step's status. ✅ done, 🔄 in progress, ⬜ not started. Sourced from PROGRESS.md, so it's accurate even across sessions. Shows exactly where you are and what comes next.

progress view status tracker visual
/stuck
I'm Stuck

Climbs the Socratic Ladder. Ask → nudge → skeleton → reveal. Never jumps straight to the answer. Ends by asking you to explain what you just learned.

socratic ladder nudge skeleton reveal

Adapts to Any Dataset.

Sensei reads your column names and values, detects the domain, and tailors every example and analogy to match. No generic explanations. No forced analogies.

💰
Fintech
Finance & Banking
  • Fraud detection
  • Credit risk scoring
  • Loan eligibility
  • Transaction anomalies
🏥
Healthcare
Clinical & Medical
  • Disease prediction
  • Readmission risk
  • Lab value analysis
  • Diagnosis classification
🛒
Ecommerce
Retail & Commerce
  • Customer churn
  • Recommendation systems
  • Customer LTV
  • Cart abandonment
🎮
Gaming
Games & Apps
  • Player churn
  • Whale detection
  • Session analysis
  • Level progression

Two Ways to Install.

Pick whichever fits your workflow. Both take under two minutes.

bash
# One-time: install the 6 slash commands globally
mkdir -p ~/.claude/commands
curl -fsSL https://raw.githubusercontent.com/karywnl/sensei/main/.claude/commands/sensei.md -o ~/.claude/commands/sensei.md
curl -fsSL https://raw.githubusercontent.com/karywnl/sensei/main/.claude/commands/start.md -o ~/.claude/commands/start.md
curl -fsSL https://raw.githubusercontent.com/karywnl/sensei/main/.claude/commands/next.md -o ~/.claude/commands/next.md
curl -fsSL https://raw.githubusercontent.com/karywnl/sensei/main/.claude/commands/checkpoint.md -o ~/.claude/commands/checkpoint.md
curl -fsSL https://raw.githubusercontent.com/karywnl/sensei/main/.claude/commands/roadmap.md -o ~/.claude/commands/roadmap.md
curl -fsSL https://raw.githubusercontent.com/karywnl/sensei/main/.claude/commands/stuck.md -o ~/.claude/commands/stuck.md
# In any folder you want to learn DS in:
cd your-ds-project/
# Open Claude Code and run:
/sensei # bootstrap this folder as a classroom
/start # begin your learning session
💡 Installs the 6 slash commands globally — one-time setup. Then in any folder you want to learn DS in, run /sensei to turn it into a classroom.