How to use Rows AI

Learn how to use AI inside the spreadsheet to automate tasks like summarization, classification, or translation.

Rows AI Template

There's a near-infinite amount of tasks you can solve using Rows AI. Use this template showcase to get started with +10 pre-built examples.

About Rows

Rows is the easiest way to import, transform, and share data in a spreadsheet. It combines a spreadsheet editor, +50 integrations with the tools you use every day, a powerful AI Analyst✨, and a sharing experience to instantly turn any spreadsheet into a web app, a form, or a dashboard.

Product image

Connect the Rows AI integration

To connect the integration, open a new spreadsheet, and search for the Rows AI inside the Data panel.

SCR-20250808-jsmj-2

Finally, Connect the integration to get started.

Using the Rows AI functions

The Rows AI integration comes with proprietary functions that automate prompts to address specific types of tasks:

  • ASK_AI: ask anything to Rows AI via the standard prompt
  • CLASSIFY_AI: classify any text into user-defined categories
  • APPLY_TASK_AI: apply a task to a text or data
  • FIND_FACT_AI: find fact about a subject
  • EXTRACT_AI: extract any concepts or portions from a text
  • SUMMARIZE_AI: summarize a text employing a specific style
  • SENTIMENT_ANALYSIS_AI: run sentiment analysis on a text
  • REWRITE_AI: rewrite a text based on given instructions
  • AD_CONTENT_AI: create ad content based on your product, tone of voice, and channel
  • META_DESCRIPTION_AI: create meta descriptions based on a page title and specific keywords
  • PRODUCT_TITLE_AI: create product titles from a product description
  • PRODUCT_DESCRIPTION_AI: create a product description based on a set of features
  • KEYWORDS_AI: create keywords for your SEO/Ads campaigns
  • KEYWORDS_MISSPELLING_AI: generate the most common misspelling of a keyword

You can use them via the Autocomplete in the editor,

SCR-20250808-jsmj-2

Examples of Rows AI in action

There are many things you can do with AI in Rows:

  • Clean Up Data: Capitalize text, remove whitespaces, and unnecessary text, parse email domains, and more.
  • Text Classification: Tag emails, classify support tickets or product feedback.
  • Sentiment Analysis: Determine the overall sentiment of a piece of text, as well as identify specific opinions and emotions within the text. Useful for product reviews, feedback, or social media comments.
  • Extraction: Extract concepts or portions from a text.
  • Summarization: Condense news articles, messages, legal documents, research papers, and more, according to a specific style.
  • Content generation: Generate content for advertising purposes, such as keywords, headlines, subheadlines, or body copies.
  • Translation: Translation texts into other languages.
  • Find facts: answer facts about a specific subject.
  • Create Tables with data: Create tables with dummy or public data.
  • Find and Enrich data: Generate dummy data or public data points for things you already have on your spreadsheet, like countries and companies.

Let's go through each of them.

Extract details from addresses

Goal

Extract the Zip Code and State from an address.

Examples

1=EXTRACT_AI("Zip code",A2)
2=EXTRACT_AI("State",A2)

Details

Assumes that A2 contains the address to clean up.

Extract feature requests from user feedback

Goal

Extract feature requests from a list of cohesive user feedback.

Examples

1=EXTRACT_AI("feature request",A2)

Details

Assumes that A2 contains the feedback to inspect.

Extract contact details from emails

Goal

Extract phone number from email.

Examples

1=EXTRACT_AI("phone number",A2)

Details

Assumes that A2 contains the email text.

Classify job titles by seniority

Goal:

Classify a Job Title according to its seniority - e.g. C-Level, Junior, Senior

Example:

1=CLASSIFY_AI(A2, "C-Level, Senior, Mid, Junior, Intern",false)

Details:

  • Add the job title to be classified as a first argument (here, cell A2)
  • Add all the categories you want to classify your titles into as a second argument
  • Set multi-tag as false to get a strict single-tag categorization
  • Leave the temperature blank (default 0) to have well-defined answers

Classify customer tickets by type

Goal:

Classify a list of cohesive customer tickets by type, e.g. engineering, sales or billing

Example:

1=CLASSIFY_AI(A2, "Engineering, billing, sales",false)

Details:

  • Add the customer tickets to be classified as a first argument (here, cell A2)
  • Add all the categories you want to classify your tickets into as a second argument
  • Set multi-tag as false to get a strict single-tag categorization
  • Leave the temperature blank (default 0) to have well-defined answers

Run sentiment analysis

Goal

Analyze comments from social media accounts and extract the sentiment from the text. While this can be performed through the CLASSIFY_AI() function, through the right categories, we introduced the SENTIMENT_ANALYSIS_AI to make the task easier to solve.

Example:

1=SENTIMENT_ANALYSIS_AI(A2)

Details:

Add the comment to classify as a first argument (here, cell A2)

Clean up company names

Goal:

Clean up a list of company names by removing legal abbreviations.

Example:

1=APPLY_TASK_AI("Remove legal entity abbreviations like GmbH, LLC, Inc., emojis, special characters and unnecessary text",A2)

Details:

Assumes that A2 contains the company name.

Translate product reviews from customers

Goal

Translate product reviews done by customers in third-party platforms - e.g. App Store, Trustpilot - to English.

Example

1=TRANSLATE_AI(A2,"english",1)

Details:

  • The function assumes that A2 contains the product review text
  • Sets a temperature of 1, to help the engine recognize nuances and idiomatic expressions

Summarize product reviews from customers

Goal:

Summarize the product review into two bullet points.

Example:

1=SUMMARIZE_AI(A2,"two bullet points")

Details:

  • The function assumes that A2 contains the product review to be summarized
  • Add the style of the summary as the second parameter, e.g. "two bullet points", or "one-liner"

Rewrite a text

Goal:

Rewrite a text to make it sound more native or change its style.

Example:

1=REWRITE_AI(A2,"add emoji")

Details:

  • The function assumes that A2 contains the text to be rewritten
  • Add the instructions for the rewrite as the second parameter, e.g. "add emoji", or "speak like Shakespeare"

Generate keywords

Goal: Generate keyword ideas based on a topic.

Example:

1=KEYWORDS_AI(A2,5,"informational")

Details: Assumes that A2 contains the topic.

💡 Change the type parameter to 'Commercial' or 'Transactional' to generate keywords oriented to purchase.

1=KEYWORDS_AI(A2,5,"commercial")

Generate meta descriptions

Goal:

Generate a meta description based on a page title and keywords.

Example:

1=META_DESCRIPTION_AI(A2,B2)

Details:

  • The function assumes that A2 contains the page title
  • The function assumes that B2 contains the keywords

Generate product titles

Goal:

Generate a number of product titles based on a product description.

Example:

1=PRODUCT_TITLE_AI(A2,,3)

Details:

  • The function assumes that A2 contains the product description

Generate product descriptions

Goal:

Generate product descriptions based on a list of features.

Example:

1=PRODUCT_DESCRIPTION_AI(A2)

Details:

  • The function assumes that A2 contains the list of features to be included in the description

Create a table with dummy names and email addresses

Goal:

Create a table with 5 dummy names and email addresses to test a feature in QA. While this can be performed through a specific-enough prompt within the ASK_AI() function, we introduced the CREATE_LIST_AI to make the task easier to solve.

Example:

1=CREATE_LIST_AI("Full names and email address",5,,500)

Use our Actions wizard to simply the table creation. Look for 'Create list' in our Actions menu, and fill all parameters as follows:

Screenshot 2023-03-09 at 10.40.53

Once clicked on 'Next Step', the table will show up below:

Screenshot 2023-03-09 at 10.41.09

Details:

  • 💡 Define a value for the max_token parameter - in this case, 500 - to allow enough tokens for the complete translation of the text.