Design relational data tables without any technical knowledge.
Tables link to each other, so a customer is written once and every order points at that customer. Click from table to table, and open the data in one click.
Demo: the Tables view of the Example Shop. A pointer clicks from Orders to its parent Customers, then to Tiers, then back to Customers. Each click moves the clicked table to the centre. Then the pointer clicks Data and the 8 customer rows show.
Fixed lists are small tables.
Status, tier and currency are tables of allowed values. A value that is not on the list is refused, on an edit and on an import.
Move from parent to child.
A row shows its children, and a new child made from the parent is linked to the parent.
Totals stay current.
A lookup counts or adds up the child rows, and the lookup runs again when a child changes.
Every change has a name and a time.
Each row records each change: what changed, when, and who made the change. That is a person, or the action that ran. Old values stay in the archive of each cell, old files included.
Demo: the page of one order in Orders. It shows the fields and the system fields, then the History: each change, the action or person that made it, and when.
Load spreadsheets through the import queue.
Send a spreadsheet to a table, by upload, from a file on a row, or as an email attachment. The import matches its headers to your columns by meaning, so "Buyer Email" fills Customer Email and "SKU" fills Product Code. When the import is not sure, it asks a person and does not guess.
Demo: the import page of a messy marketplace file in the Example Shop. The AI matched six headers to the Orders columns by meaning and rewrote two values. One question is left: how the dates are written. A pointer answers Day first and clicks Load. The Orders data screen opens: the six new rows arrive together, then the table actions fill each row with its total, customer name, tier and region.
Step 1
Send the file
Upload an Excel or CSV file, or let an action deliver the file: a file on a row, or an email attachment.
Step 2
Match the headers
A header with the same name as a column maps at once. AI maps the others, one column at a time.
Step 3
Answer what is open
Pick the header row, answer a date-format question, correct values or leave rows out.
Step 4
Load
The rows go in, and the table’s actions run on the new rows. The next file of the same layout maps the same way.
No double loads.
If a table has a unique column, a file that was already loaded stops in the queue and writes nothing.
Large files.
One sheet can hold up to 2,000,000 cells. One uploaded file can be up to 100 MB.
Every column has a type.
A type tells people, imports and actions what a column holds. A PDF column takes PDFs. A currency column takes amounts. An action that needs a spreadsheet says so in its input.
Text
Aa
Text
Aa
Aa
Telephone number
URL
Numbers
#
Integer
#
Decimal
£
Currency
Percentage
Time
Date
Time
Date and time
Date and time with zone
Year and month
Year
Files
File
Excel
Aa
Word
Image
CSV
Other
True or false
Aa
Coordinate
Reference
Find, export and connect.
What you can do with a table once the rows are in.
Filter and search
Filter on any column: equal, contains, greater than, empty, in a list. Combine filters to narrow the rows.
Export to Excel
Download what you see, with your filter and sort, up to 100,000 rows. For a CSV or a Word file, use a generate action.
Status on every row
Six system columns come with every table. Action Status shows how the row’s actions went.
Rows from other systems
Another system can create rows, one at a time or in bulk, with a workspace API key.
Access by workspace
Members are admins, writers or members, and an admin can make members read-only. Access applies to the whole workspace.
Security and compliance.
ISO 27001, SOC 2 Type 2 and GDPR are in progress. Oneleet shows the live status of each certification.
The agentic system builder
© 2026 morphcloud