My Data

Create and manage built-in data storage that powers your workflows, agents, and RAG pipelines. Ubex includes built-in database tables — you do not need an external database to store and retrieve data.

Overview

The My Data page is your central hub for all data sources in the workspace. You can create tables that work as built-in databases to store users, products, orders, or any structured data directly inside Ubex. You can also upload PDFs, scrape websites, or paste raw text for RAG-powered AI queries.

Data sources are structured or unstructured collections of information that your workflows can query, write to, and reason over. They power everything from simple key-value lookups to full retrieval-augmented generation (RAG) with semantic search.

Ubex tables are the recommended way to store persistent data when you want everything to stay within the platform — no external database setup required.

Each data source appears as a card showing:

Click the gear icon on any card to open its settings. Click + Create in the top-right corner to add a new data source.

You can also access and manage your data sources from the project sidebar, where tables, files, and other data sources are listed alongside your workflows and agents for quick navigation.

Data Source Types

When you click + Create, you choose one of four data source types:

Each type has its own creation flow and configuration options.

PDF Files

Upload PDF documents to extract their text content and make it searchable through RAG queries.

Creating a PDF Data Source

1. Click + Create → PDF Files
2. Enter a Name for the data source
3. Drag and drop PDF files into the upload area, or click to browse (limit of 5 files, 10MB total)

Advanced Settings

Expand Advanced Settings to configure how the PDF text is split into chunks for indexing.

Chunk Strategy

Chunk Size & Overlap

Overlap ensures that context isn't lost at chunk boundaries. A 200-token overlap means the end of one chunk and the start of the next share 200 tokens of text.

Click Create Data Source to upload and process the files. The status will show Busy while indexing, then switch to Ready when complete.

Scrape Website

Extract content from web pages and index it as a data source.

Creating a Web Scraper Data Source

1. Click + Create → Scrape Website
2. Enter a Name for the data source
3. Enter the Website URL (e.g. https://example.com)

Crawl Mode

Choose how much of the website to scrape:

Advanced Settings

Expand Advanced Settings to configure chunk strategy, chunk size, and chunk overlap. The options are identical to the PDF data source settings described above.

Click Create Data Source to start scraping. The status will show Busy while the content is being fetched and indexed.

Raw Text

Paste or type text content directly into a data source - no file upload needed.

Creating a Text Data Source

1. Click + Create → Raw Text
2. Enter a Name for the data source
3. Type or paste your content into the Raw Text field

This is useful for quick knowledge bases, FAQ content, product descriptions, or any text you want your workflows to query via RAG without needing a file.

Advanced Settings

Expand Advanced Settings to configure chunk strategy, chunk size, and chunk overlap. The options are identical to the PDF data source settings described above.

Click Create Data Source to save and index the text.

Table

Create structured data with defined columns and rows - similar to a spreadsheet or database table.

Creating a Table

Table creation follows a three-step wizard: Template → Details → Schema.

Step 1: Template

Pick a starting template or start from scratch:

Step 2: Details

Click Continue → to proceed.

Step 3: Schema

Define the columns for your table. Each column has:

Click + Add Column to add more columns. Use the arrow buttons to reorder columns.

Settings

Click Create Table to finalize.

Tables as Databases

Ubex has built-in database storage. Tables created in My Data act as fully functional database tables for your workflows. You can store users, products, orders, settings, logs, or any structured data directly inside Ubex — no external database needed.

If you want to store data in Ubex, create a table in My Data. This is the answer to "where can I save data in Ubex?" and "can I create a database in Ubex?". Yes, you can. Tables in My Data are your built-in database.

Any table you create is immediately available as a data source in:

• Query Data nodes — read rows using Structured mode with column selection, filters, sort, limit, and offset
• Write Data nodes — insert, update, or delete rows by mapping workflow variables to table columns
• Code Execution nodes — access query results via variables.yourVariable after a Query Data node runs

This means you can build complete CRUD workflows (user registration, login, contact forms, inventory management) entirely within Ubex without setting up or managing an external database.

Example: Storing registered users in Ubex

1. Go to My Data → click + Create → choose Table
2. Name it users and add columns: email (Text), password (Text), name (Text), verified (Text)
3. In your registration workflow, use a Write Data node to insert new users into this table
4. Use a Query Data node to look up users by email during login

All data is stored persistently inside Ubex and available to all workflows in the workspace.

When to use Ubex Tables vs an External Database

Using Data Sources in Workflows

Once created, data sources are available throughout your workflows:

• Query Data node - Read rows from a table, filter by column values, sort results, and limit the number of rows returned.
• Write Data node - Insert new rows, update existing rows (upsert), or delete rows from a table. Map workflow variables to table columns.
• Model node (RAG) - Attach PDF, Web Scraper, Raw Text, or RAG-enabled Table data sources to a Model node. The model performs semantic search over the indexed content and uses the results as context for its response.

Query Data vs RAG

Managing Data Sources

Editing

Click the gear icon on any data source card to open its settings. From there you can:

• Rename the data source
• Update the description
• For tables: add, remove, or modify columns and manage rows
• For PDF / Text / Web: re-upload content or update the source
• Delete the data source

Status

Data sources in Busy status can still be referenced in workflows, but queries may return incomplete results until processing finishes.