Tomat Documentation
Tomat HomeCommunity
  • 👋Introduction to Tomat
  • Getting Started
    • Product Updates
    • Beginner's Guide
    • FAQ
  • Product overview
    • Home Page
    • Exploring Data
      • Data Catalog
      • Exploring Datasets
      • Statistics Panel
    • Designing Flows
      • Creating Flows
      • Flow Designer Guide
        • Working with Canvas
        • Using Groups
        • Working with Table
      • Managing Flows
      • Demo: Building a Simple Flow
    • Executing Flows
      • Running Flows
      • Jobs overview
    • Building Reports
      • Designing Reports
      • Running Reports
      • Reports Page
    • Connectors
      • Data Storages
      • Data Sources
      • Enrichments
    • Tomat Use Cases
  • Data Transformation
    • Transforms
      • Source
      • New Empty Table
      • Output
      • Chart
      • New Column
      • If...Then
      • Rolling Functions
      • Column Type
      • Columns Edit
      • Sort
      • Filter
      • Remove Duplicates
      • Split Text
      • Extract Text
      • Find and Replace Text
      • Match Text
      • Join
      • Union
      • Group By
      • Pivot
      • Unpivot
      • To JSON
      • From JSON
      • API Call
      • AI Column
      • AI Table
    • Formulas
      • What are Formulas?
      • Math Functions
        • Abs
        • Ceiling
        • Exp
        • Floor
        • IsEven
        • IsOdd
        • Ln
        • Log
        • Log10
        • Mod
        • Pi
        • Power
        • Quotient
        • Round
        • RoundDown
        • RoundUp
        • Sign
        • Sqrt
        • Truncate
      • Trigonometric Functions
        • Acos
        • Asin
        • Atan
        • Atan2
        • Cos
        • Cot
        • Degrees
        • Radians
        • Sin
        • Tan
      • String Functions
        • Compare
        • Concat
        • Contains
        • In
        • CountMatches
        • CountMatchesRegexp
        • EndsWith
        • EndsWithRegexp
        • Extract
        • FindMatchOfString
        • FindMatchOfRegexp
        • FindMatchesOfString
        • FindMatchesOfRegexp
        • Left
        • Length
        • Lower
        • Matches
        • Pad
        • ProperCase
        • RemoveSymbols
        • RemoveWhitespaces
        • Repeat
        • Replace
        • ReplaceRegexp
        • Reverse
        • Right
        • Spaces
        • Split
        • SplitRegexp
        • StartsWith
        • StartsWithRegexp
        • Stuff
        • Substring
        • SubstringDelimiter
        • SubstringRegexpDelimiter
        • Trim
        • Upper
      • Date & Time Functions
        • Date
        • DateAdd
        • DateAdd
        • DateDiff
        • DateDiff2
        • DateFromParts
        • DateTime
        • DateTimeFromParts
        • DateTrunc
        • DayName
        • DayOfMonth
        • DayOfWeek
        • DayOfYear
        • Hour
        • Minute
        • Month
        • MonthName
        • Now
        • Quarter
        • Second
        • Time
        • TimeFromParts
        • Today
        • Week
        • Year
      • Aggregate Functions
        • Array
        • ArrayIf
        • Avg
        • AvgIf
        • Count
        • CountA
        • CountIf
        • CountUnique
        • Max
        • MaxIf
        • Median
        • MedianIf
        • Min
        • MinIf
        • Mode
        • ModeIf
        • Percentile
        • Quartile
        • StdDev
        • StdDevIf
        • Sum
        • SumIf
        • SumProduct
        • Variance
        • VarianceIf
      • Conversion Functions
        • ToArray
        • ToBoolean
        • ToDate
        • ToDateTime
        • ToDecimal
        • ToInteger
        • ToObject
        • ToTime
        • ToString
      • Misc Functions
        • At
        • IsMissing
        • RowNumber
        • Random
        • If
        • Coalesce
        • True
        • False
        • Null
        • $target
      • Window Functions
      • Custom Functions
      • Data Types
      • Supported Date Parts
      • Regex: List of Tokes
  • PRICING & BILLING
    • Plans, Subscriptions, and Credits
    • Tomat for Education
  • Integrations
    • Data Storages
      • Snowflake
      • PostgreSQL
  • Tutorials
    • Merge Columns
    • Join Types
    • Union Introduction
    • Window Functions
    • What is Unpivot?
    • JSON Format Tutorial
    • Using Regex
Powered by GitBook
On this page
  • Overview
  • Settings
  • What to ask
  1. Data Transformation
  2. Transforms

AI Column

Adds a new column based on GPT prompt

Last updated 9 months ago

Overview

By adding this node to your canvas, you can simply input a natural language prompt to automatically generate a new column. Whether you're looking to enrich existing data, extract specific information, or format values, the AI Column Node acts as a 'black box' that simplifies these processes into a single, user-friendly step.

Pay attention, that if you use this node with large datasets, you can quickly reach the limits of GPT tokens.

Settings

Write your prompt in natural language to add a new column based on your request. You should mention at least one of the existing columns (use @ symbol), based on which AI will return new values.

AI knows data in the mentioned columns, their names, and types.

Use our prompt library for ideas, don't forget to replace 'Column' with your column name.

Button "One row"

You can fill a new column row by row here. We recommend always using this option first to be sure you get the expected result.

Button "All rows"

If you are comfortable with one-by-one calling, you can fill all rows simultaneously. It will not refill rows already containing the result.

Button "Remove answers"

If you get some errors in the answer or want to change the request - use this button to remove results for all rows.

AI will return a name for a new column based on the context of the query. You are free to change it later.

An alternative way to get the result for a specific row is to find a newly created column in the table preview and click the "Generate row" link.

What to ask

  • Any manipulation with data in the column(s): extract, enrich, replace, translate, format, etc.

  • AI Column doesn’t work without a column mentioning.