Your first upload

Before you start

This guide shows you how to turn your first file into a table. Every upload needs an existing destination dataset. If your workspace is empty, create a dataset first; Loam does not pick a destination for you.

Steps

1. Open the dataset that should contain the table and choose Add table. If it does not exist, create it with the ordinary New dataset action first.
2. Choose or drag in a CSV file. You'll see its name, size, and inferred columns.
3. Check the selected dataset and table name. The destination is the same dataset shown in the sidebar and used in the table's SQL path.
4. Select Upload and create Table. You'll see Uploading with progress, followed by Validating and committing….
5. When the table path, column count, and row count appear, select View table.

CSV creates a Table from its inferred schema. SQLite uploads show their tables and columns so you can select one source table before Loam creates the destination. The same dialog can append, upsert, or replace data in an existing Table from CSV, TSV, JSON, Excel (.xlsx), or Parquet files whose columns match the target Table.

What to expect

When processing finishes, Loam shows the table path, column count, and row count. Select View table to open your data with its column names and types.

From here you can:

• Open the SQL Console to query your data
• Use Explore to browse and filter visually
• Connect your AI and ask questions in plain English

Troubleshooting

If Loam does not accept your first file, check that it is a CSV no larger than 100 MB. If the table name is rejected, use lowercase letters, digits, and underscores, start with a letter, and keep it between 1 and 64 characters. Then try again.