> ## Documentation Index
> Fetch the complete documentation index at: https://docs.clickml.app/llms.txt
> Use this file to discover all available pages before exploring further.

# Quickstart guide to building ML pipelines

> Build your first end-to-end machine learning pipeline in ClickML Studio in under five minutes by connecting data, model, and evaluation components.

This guide walks you through a minimal end-to-end pipeline: load a CSV, split the data, train a Random Forest, and evaluate it.

## 1. Create a project

1. Sign in at [app.clickml.com](https://app.clickml.com).
2. Click **New Project**.
3. Give it a name, select a problem type (**Classification** or **Regression**), and confirm.

You land on the canvas — a blank workspace with a toolbox on the left.

## 2. Add a CSV Reader

Drag the **CSV Reader** component onto the canvas.

Click it to open the inspector panel. Upload your CSV file (or paste a URL). Make sure your file has a header row.

## 3. Add a Train-Test Split

Drag **Train-Test Split** onto the canvas, to the right of CSV Reader.

Connect the CSV Reader's output handle to the Train-Test Split's input handle by dragging between the two dots.

In the inspector, set your **target column** (the column you want to predict) and adjust the test size if needed (default: 20%).

## 4. Add a model

Drag **Random Forest Classifier** (or **Random Forest Regressor** for regression problems) onto the canvas.

Connect the `X Train` and `Y Train` output handles from Train-Test Split to the model's input.

## 5. Add Inference

Drag the **Inference** component onto the canvas.

Connect the model output and the `X Test` handle from Train-Test Split to Inference's inputs.

## 6. Add Evaluation

Drag **Evaluation** onto the canvas.

Connect `Y Test` from Train-Test Split and the predictions output from Inference to Evaluation's inputs.

Set the **problem type** in Evaluation's inspector to match your project.

## 7. Run

Click **Run** in the top bar. The pipeline executes and each component shows a status indicator.

Open Evaluation to see your metrics: accuracy, F1, confusion matrix (classification) or MAE, RMSE, R² (regression).

***

## Minimal pipeline diagram

```
CSV Reader → Train-Test Split → Random Forest → Inference → Evaluation
```

From here you can add preprocessing steps (Scaling, Encoding, Missing Values) between the CSV Reader and the split, or swap in a different model to compare results.