> AI agents: this is one page from Mammoth Analytics documentation. The index of all pages as Markdown is https://docs.mammoth.io/llms.txt. Append `.md` to any docs URL, or send `Accept: text/markdown`, to get Markdown.

# Quick start guide

> Build your first Mammoth pipeline with this hands-on tutorial: upload, explore, transform, and export a messy sales file.

This hands-on tutorial is for new Mammoth users. It walks you through the core workflow (upload, explore, transform, and export) using a small sales file, and ends with a clean, exportable dataset.

**What you'll accomplish**: Transform messy sales data with duplicates and inconsistent formatting into a clean dataset ready for analysis.

**Prerequisites**: A Mammoth account ([sign up for Mammoth](https://mammoth.io/signup))

**Data**: A CSV file of sales orders with messy product names, mixed date formats, and duplicate rows (described in Step 1)

![Diagram of a pipeline: raw data passes through Filter, Standardize, and Calculate tasks to produce clean data](https://docs.mammoth.io/api/v1/images/20260324_061100_Completepipelineworkflowdiagram.jpg)

## Step 1: Upload the sales CSV as a Dataset

This step gets the sales data into Mammoth.

1. Prepare a CSV file of sales orders, such as `sales_data_messy.csv`, with columns `product_name` and `order_date`:

   - Product names written in several variations

   - Dates in inconsistent formats

   - Duplicate entries

2. In Mammoth, click **Data** in the left sidebar to open the Data Library.

3. Click the **New Dataset** button at the top right of the Data Library.

4. In the New Dataset dialog, choose **Upload**

5. Drag your file into the drop zone and click **Submit**

![Animation of the empty Data Library: clicking New Dataset, choosing Upload, and adding a file](https://docs.mammoth.io/api/v1/images/20260302_152934_upload-file.gif)

**What just happened**: Mammoth imported your CSV and created a Dataset (a standardized data container). The data is now stored in your Project. Open the new dataset from the Data Library list to see it in a View. A View is your interactive workspace for exploring and transforming data.

![Data Library showing sales_data_messy.csv with 7 columns and 518 rows, a Preview, and an Upload complete message](https://docs.mammoth.io/api/v1/images/20260408_155738_File_uploaded_data-lib.jpg)

## Step 2: Explore the sales data with Explore Cards

Before transforming data, check its quality and patterns. Explore Cards profile each column automatically.

### Profile the sales columns with Explore Cards

When your View opens, **Explore Cards** appear above the data grid. They are automatic visualizations showing the distribution of values in each column.

1. Click on the **product_name** column's Explore Card

   - See the bar chart showing value distribution
   - Notice inconsistencies: "iPhone 13", "iphone 13", "IPHONE 13" all appear separately
   - This reveals a standardization opportunity

2. Click on the **order_date** column's Explore Card

   - See the date distribution histogram
   - Notice format variations that need cleaning

3. Open **Data Quality** in the footer below the data grid (when it is enabled for your workspace)

   - Shows the quality score, column completeness, critical issues, and recommendations
   - Click **View Details** for the full analysis

![View of sales_data_messy.csv with Explore Cards for product_name, order_date, unit_price, and region above the data grid](https://docs.mammoth.io/api/v1/images/20260408_155758_VIewwithcards.jpg)

**What just happened**: Explore Cards automatically profiled every column, so you can spot quality issues without manual analysis.

Double-click any value in an Explore Card to filter your data to just those rows. This creates a filter "pill" that you can remove at any time or convert into a permanent pipeline task.

### Data quality issues in the sales file

The sample data contains common real-world issues:

- **Product names**: Multiple variations ("iPhone 13", "iphone 13", "IPHONE 13")
- **Duplicate records**: Same order appearing multiple times
- **Date filtering needs**: Mix of old and recent orders

Step 3 fixes these problems.

## Step 3: Clean the sales data with three tasks

This step cleans and standardizes the data using visual transformations. Each transformation creates a Task in your Pipeline, which can be applied to future data.

### A. Standardize product names

1. Click the three-dot menu (⋮) on the **product_name** column header

2. Select **Bulk Replace** from the menu

3. Mammoth analyzes the column and suggests groupings of similar values:

   - "iPhone 13", "iphone 13", "IPHONE 13" → "iPhone 13"
   - Similar patterns for other products

4. Review the suggestions and click **Apply**

5. Notice the Pipeline panel on the left now shows "**Task 1: Bulk Replace**"

![Bulk Replace panel on product_name with Group Suggestions listing AirPods Pro, IPAD AIR, and an iPhone 13 group of 3 values](https://docs.mammoth.io/api/v1/images/20260408_155818_Bulk_relplace_config.jpg)

![Bulk Replace review step showing the iPhone 13 group with Previous and Apply buttons](https://docs.mammoth.io/api/v1/images/20260408_155831_Bulk_replace_apply.jpg)

![View with a Bulk Replace task on product_name in the Pipeline panel and standardized iPhone 13 values](https://docs.mammoth.io/api/v1/images/20260408_155850_Pipeline_with_bulk_replace.jpg)

**What just happened**: You fixed inconsistent product names. The Bulk Replace function identified similar values and grouped them. This same logic now applies automatically to any new data you add to this Dataset.

### B. Remove duplicate rows

1. Click the **+** button (tooltip **Add a task**)

2. In the task picker, open **Filter Label & Replace** and select **Remove Duplicate Rows**

3. Under **Evaluate duplicates from**, choose which columns to compare (leave the default to compare all columns)

4. Click **Apply**

5. Notice the Pipeline shows "**Task 2: Remove Duplicate Rows**" and the row count decreased

![Task picker opened from the plus button in the Pipeline panel, with the search box, the category list and the Filter, Label & Replace category expanded to show Conditional Filter and Remove Duplicate Rows](https://docs.mammoth.io/api/v1/images/20261009_092921_Task_picker_filter_label_replace.jpg)

Tasks execute from top to bottom in your Pipeline. You can drag tasks to reorder them, and Mammoth will re-execute the entire pipeline in the new sequence.

### C. Filter to recent orders

1. Click **+** → **Filter Label & Replace** → **Conditional Filter**

2. Configure the filter:

   - Select column: **order_date**
   - Choose condition: **is later than**
   - Set date: **2023-01-01**

3. Click **Apply**

4. The Pipeline now shows "**Task 3: Conditional Filter**"

![View with three tasks in the Pipeline panel: Bulk Replace, Remove Duplicate Rows, and Conditional Filter, leaving 349 rows](https://docs.mammoth.io/api/v1/images/20260408_155908_three-task-in-pipeline.jpg)

**What just happened**: You built a three-task transformation pipeline without writing any code. Each task executes in sequence, and you can see the results immediately. The pipeline is saved, so when new data arrives, these same transformations apply automatically.

## Step 4: Review the three-task pipeline

This step shows how to check each task in the pipeline.

### Inspect each task's transformation logic

1. Look at the **Pipeline panel** on the left sidebar

2. Click on any Task to see its configuration:

   - Shows exactly what parameters you set
   - Displays the logic applied to your data

3. Click **Preview** on each task to see before/after data:

   - View data state at any point in the pipeline
   - Understand how each transformation affects your results

4. Notice the complete audit trail:

   - Every change is listed as a Task
   - Team members can see and understand your logic

![Previewing data at Task 2, Remove Duplicate Rows, with the Exit Preview bar and the task's configuration panel open](https://docs.mammoth.io/api/v1/images/20260408_160139_Preview_task_with_remove_duplicate_config.jpg)

**What just happened**: Your manual work is now a repeatable pipeline. When new sales data arrives, these same transformations execute automatically.

You can go back to any pipeline task and see the data state at that point, which helps when you need to trace how a result was produced.

### What the three-task pipeline demonstrates

This pipeline shows three properties of Mammoth pipelines:

**Automation**: Future data batches can run through the same tasks

**Auditability**: Every transformation is listed in the Pipeline

**Reusability**: This pipeline can be copied and adapted for similar datasets

## Step 5: Export the cleaned data as CSV

This step downloads the cleaned data.

1. Click the download icon in the top toolbar to open the **Save file as** menu

2. Choose your export format:

   - **CSV**: Most common, works everywhere
   - **XLSX**: An Excel workbook
   - **Mammoth Dataset**: Save your data to a new dataset

3. For this tutorial, select **CSV**

4. Click **Confirm**

5. Open the file—notice clean, standardized data:

   - Product names consistent
   - Duplicates removed
   - Only recent orders included

![Save file as menu from the download button, with CSV selected, XLSX, Mammoth Dataset, and a Confirm button](https://docs.mammoth.io/api/v1/images/20260408_160219_Download-options.jpg)

**What just happened**: You transformed messy data into clean output. The pipeline you created is saved and reusable.

### Alternative next steps

Instead of downloading, you could:

- **Schedule automatic execution** so fresh data processes every night
- **Create a dashboard** using the AI-powered dashboard creator

## Summary of the quick start workflow

You have now run the complete Mammoth workflow:

1. **Upload**: Drag-and-drop or file selection gets data into Mammoth
2. **Explore**: Explore Cards profile each column without manual work
3. **Transform**: Point-and-click functions, with no code or formulas
4. **Reuse**: A pipeline applies to new data automatically
5. **Export**: Multiple formats and destinations are available

The pipeline gives you reusable preparation, fewer manual errors, a visible audit trail of every task, and self-service without IT dependencies.

## Next steps

Choose a learning path based on your role:

### For business analysts

**Recommended path**: [Getting started as a data analyst](https://docs.mammoth.io/learn/getting-started/as-a-data-analyst)

Focus on:

- Automating monthly reporting workflows
- Building executive dashboards
- Integrating with Excel and PowerBI
- Scheduling pipelines for automatic execution

### For technical users

**Recommended path**: [Getting started as a technical user](https://docs.mammoth.io/learn/getting-started/as-a-technical-user)

Focus on:

- Advanced transformation functions
- SQL Query capabilities
- API integration and automation
- Team enablement and template creation

### Deep dive on Mammoth concepts

To understand how the platform is organized, read [Core concepts](https://docs.mammoth.io/learn/getting-started/core-concepts).

Learn about:

- Workspace, Project, Dataset, Batch, View hierarchy
- How Pipelines and Tasks work
- Live Connections for database integration
- Data versioning and audit trails

### Explore specific Mammoth features

**Explore Cards**: [Explore Cards](https://docs.mammoth.io/learn/explore-discover/explore-cards-overview)\
**Transformations**: [Transform data overview](https://docs.mammoth.io/learn/transform/)\
**Dashboards**: [Dashboards](https://docs.mammoth.io/learn/dashboard/)\
**Troubleshooting**: [Troubleshooting common issues](https://docs.mammoth.io/learn/troubleshooting/)

### Try the workflow with your own data

1. Upload a real file from your work
2. Explore it using Explore Cards
3. Build transformations for your actual business problems
4. Share results with your team

**Tip**: Start with a small dataset (< 10,000 rows) to learn quickly, then apply those skills to larger production datasets.

## Need help?

- **In-app documentation**: Click the 📘 book icon next to a transformation to open its related documentation and learn how that feature works.
- **In-app Support**: Use the Support option in the sidebar to ask questions, report issues, or share feedback.

**Remember**: You now have the foundation to automate your own data workflows.

---
Source: https://docs.mammoth.io/learn/getting-started/quick-start.md · Updated: 2026-10-13