All posts
Published at Tue Jun 13 2023 in
Rows HQ

Introducing the AI Analyst

Henrique Cruz
Henrique Cruz, Head of Growth, Rows
Thumbnail AI Analyst

Today marks a big milestone in Rows' journey, as we introduce the AI Analyst. It comes with a simple vision: to give one billion spreadsheet users the capabilities of a data analyst.

This launch is the natural next step to define the future of spreadsheets. 

First, we built integrations that made it easy to automatically import data from virtually any source in a few clicks. We started with marketing tools like Google Analytics and Facebook Ads, expanded to include CRMs such as HubSpot and Salesforce, and recently added data warehouses and databases like Google BigQuery or PostgreSQL. 

Then, we made people proud to share their spreadsheets. Live sharing made spreadsheets beautiful and easy to use, even on mobile. Next came Embeds, which transformed how people shared spreadsheets across apps. Sharing your spreadsheets became seamless, interactive and fun.

Today, we are thrilled to tackle the final piece of the puzzle: transforming and analyzing data.

Why build an Analyst?

We have often said that if you know how to do a VLOOKUP, you’ll know how to use Rows (and love it). But here’s the thing: out of the billion spreadsheet users, many - perhaps most - don’t know how to do a VLOOKUP. Formulas are elegant and deterministic, but they can also be confusing and hard to use.

That’s why back in 2021 we started taking on this problem by shipping the wizard, a guided journey to create and edit spreadsheet formulas. It was a substantial leap forward, but we knew the problem wasn’t solved yet. It still takes effort to load and think about which analysis you want to do.

The advent of LLMs (Large Language Models) spurred us to rethink this issue. Our mission was clear: empower a billion people with the capabilities of a data analyst.

Why a Data Analyst? Let’s look at an actual job description of one. 

Data Analyst job descriptionA real job description of a Data Analyst

A Data Analyst blends interpretation, analysis, and reporting.

These are the same tasks we see everyday watching people use spreadsheets: they have a goal in mind (e.g. to build a sales report), they have the data, but going from the data to the end goal takes too long. Most of the time is not spent on thinking about the right way to approach the problem, but on writing formulas, filtering and sorting data, building pivot tables or just finding someone to help them analyze the data.  

Now is the right time to change that. The power of AI and the versatility of a spreadsheet are the perfect combination to solve that problem and Rows is the natural tool for the job.

This was the inspiration behind our AI Analyst. It's not just a co-pilot to help you navigate; it's a colleague that can take over the wheel, work alongside you and suggest points of view you might not have thought of yourself, for the lack of time or know-how.

AI Analyst mockOne of the first wireframes from the AI Analyst

Your number-crunching sidekick

AI AnalystIntroducing the AI Analyst

The AI Analyst is like a number-crunching sidekick. It summarizes the main take-aways from any dataset, runs in-depth analysis, and answers any questions you might have about your data. It’s a triad of AI super powers!

  1. Quick insights: Your new AI Analyst understands the key variables from a dataset and gives you an overview of 3-5 key takeaways. With easy access to copy the insights, *and* the formulas, you can peek under the hood and validate or extend the insights as you want.

  2. Deep dives: The AI Analyst extracts trends, generates pivots, and identifies patterns. You can add any of these deep dives as tables to your spreadsheet and continue digging in, add Charts or Embed the results somewhere else.

  3. Ask anything: Just ask. Rows AI Analyst can answer your queries, making it much simpler to find specific trends, create pivot tables and extract specific information about the dataset.

One of the areas of focus in this product release was privacy. Unlike most AI assisted products, the Rows AI Analyst does not use your table’s data for training models nor do we share that data (only metadata) to any external models. Your data is processed and crunched by us, guaranteeing that your data never leaves Rows. (2024-03-27 Update: we now send a bit of data, but really only a bit, like the header, and a small sample of up to 5 rows, and some basic stats like min and max of some numeric columns. See your privacy policy at rows.com/privacy to the up-to-date version of what we do with your data).

Our Analyst  is still in probation period (beta), so give it some time as it learns new tricks in the next weeks and months. 


We built the AI Analyst to make data analysis simpler, faster and more accessible. We can’t wait to show you what else we have in store for you.

Ready ✨?

Your analyst is waiting. Try the AI Analyst today.