Introduction

Hey, I'm Giani. Nice to meet you.

I normally don't put my full name on anywhere, but I did. So Giani, nice to meet you.

From Clipboard AI to a Broader Vision

I am the founder and CEO of this company, TapTapTap. We started working a few months ago on this, and I can show you what we started with, tell you what you're trying to do, and then show you the product, if that works.

So we started a few months ago working on a problem that A lot of us do.

Putting AI in Copy/Paste

So in this case, we put AI in your clipboard. You could copy anything by pressing Option and C. Look at the bottom right of the screen. You would capture it.

And when you want to put the information here, you don't have to manually fill it. You could just say Paste, and the AI would paste it for you.

So this is what we started with. You could take pretty much anything.

So let's say you're on this website and you want to make a table out of it. You could just go here. and hit paste.

Works Across Your Favorite Apps

It worked everywhere. It worked on LinkedIn. It worked on Google Docs. It worked on Slack.

On LinkedIn, if you tried pasting, it would write a message to the other person on your behalf.

Why We Built It

So we started putting AI in the clipboard because as software engineers ourselves, we really think that the best tools are coming to our software engineers and we wanted to give AI superpowers to everyone. And copy and paste felt like a very simple mechanism to give everyone AI.

My sister who works as a doctor, she spent a lot of time structuring information into a spreadsheet last year. And I thought if we can structure it for her with copy and paste, she can work a lot more efficiently.

So whenever it's pasting, you can always cancel it. Just hit cancel.

Building and Presenting the Demo

If you notice this presentation, this is running on the internet.

I used Cursor to make it. I'm a software engineer, so I have Cursor and I was trying to build something that useful for non-engineers primarily.

How many of you use Google Sheets?

Pretty cool. Perfect audience. Thank you.

The Google Sheets Use Case

So this is the same spreadsheet. A friend, she sent this to me around in June and she was like, do I actually have to go and copy and paste and copy and paste? Because even with classic tap, tap, tap, you would still copy and paste. So we said, no, we can improve on the process.

So that's how tap, tap, tap, on Google Sheets was born.

I'll move to the next slide first.

Live Demo: Enriching a Guest List in Sheets

So I'll show you these core capabilities, and then we can try things out on our own as well. But I'm happy doing live demos.

So this is a simple spreadsheet.

She hosted an event, and she had maybe 60 people show up to the event. She ended up with a list of LinkedIn URLs, and she was trying to learn who these people are.

Ask to Fill the Sheet

1So now you can just do this. You can just say, fill this spreadsheet, and then the AI will work with you. The AI will do the hard part of gathering the information.

It will always show you what it's thinking. So it can see the sheet. Interesting. It would normally.

So also to use the LinkedIn URL and to fill the rest. And then it should go all the way. Cool, let's see what it does.

How It Processes Rows

So it's gonna read the sheet first. That's what it's doing. It's reading the sheet first. It's trying to figure out how many rows actually have to be filled.

It read between A1 and A55, and now it's gonna start doing the work. So it's doing 11 to 20 first. It's calling out to the internet and figuring out who these people are.

So it's thinking, it found some of them, and it found the first five, and then it put it on the sheet for you. You don't have to do anything. You just say what to do, it'll do it for you.

Verifying Results and Handling Duplicates

I let it run, and then, see, it pulled in the information about Bukia Georgia. So she is the founder of Brinos, that is correct. She worked at Emerson in the past. Is that correct? Yes.

And she, what's the last one? She went to Babes Pullaya University. So she did go to that university. So it picked up all of the information.

So normally it does 10 at a time. Today it's decided to do five at a time, but you can always ask it to do more and it'll just keep working for you. So this is like a very basic example of how powerful it is.

But let's say I want more than the things that are on the sheet.

I'll let it run this run and then I'll stop it. So it figured out that row 24 is actually a duplicate.

So Christopher Peel is here as well. So it can just use the information that it picked previously and put the same person here.

I'll stop this for now.

Adding Emails and Phone Numbers

And I can say, can you figure out their emails and phone numbers and put it in separate columns. So you can just say this and it will go again, go through them, figure out who these people are, what their phone numbers might be, what their email addresses might be. We'll call two data providers on your behalf and we'll put the information there.

So we'll insert two columns. It's inserted some columns and then now we'll write the columns and we'll start. adding it for you.

So anything that you normally find yourself doing in spreadsheets, we want to help you with it. We want to give you an AI companion that helps you with the job.

So now it will use something called lead magic to figure out who these people are. And it will try to fetch the phone numbers and the email addresses.

Automating Sales Enrichment

So otherwise, you would maybe dump the sheet, put it into a polo, introduce it in a polo, bring it back. But we want to take care of all of it for you.

So it figured out that Lucia, she works at Oracle, so it was able to create an email address for her and it'll start putting it there.

But this is enrichment, so we've seen people use this heavily to do stuff like this. Get emails, get phone numbers, get LinkedIn. People try to figure out how many people work at a company, what the revenue is, so they use it in the sales process.

Especially if you do it on Sheets manually, you don't have to learn anything. You're already on Sheets. You can start using us almost immediately.

So I'll stop this.

Beyond Enrichment: Analysis and Modeling

Pivot Tables and Complex Models

This is another example of how you might use us. This entire demo was created by TapTapTap, and I can just ask it something like, can you create me a pivot table of salary and department? And then it'll go do that for you. You just ask it a question.

You can create fairly complex financial models as well. We had someone who was using us to model cafes and bars that he's opening in Laos. And he's just running the entire operation using us.

Like sheet native people end up finding like very interesting use cases.

This here shows you like what has changed. So I'll show you what it has added.

Versioning and Reverting Changes

You can always revert to a previous checkpoint if you don't like what it's done. And if you do that, we will go back to where it was and the pivot table should be gone. And the pivot table is gone.

Formatting, Dropdowns, and Segmentation

You can ask it to convert column I into a dropdown and it should do that.

If you're analyzing data and if you're going to categorize rows, the AI can see each and every row for you, categorize it. If you're trying to do segmentation, you can ask it to create more tabs and like pull data there.

So interesting. So it made it a dropdown. You can ask it to conditionally format stuff as well.

Planning Workflows in Sheets

Whenever we run planning, we use a spreadsheet, and this does it for us. We just put in meeting notes, and it puts it at the right places, and then we can just start working.

So these are some examples.

I can do something completely random and see what happens.

Research Workflows in Sheets

Stock Research Example

So I did some stock research prior, and this entire spreadsheet was built by TapTapTap. It went and figured out the financial data, the company overview, the SEC filings, the recent news.

The chart history is not here, but I'm happy showing a live demo, perhaps. Happy taking requests, but we can ask it to... So obviously, this is a demo.

Disclaimers and Deeper Research

You shouldn't use it for financial advice. But you can see how it will do proper research. And it might put it onto one sheet, but you can ask it to do research and create separate tabs for different kinds of information.

So you can see it's searching for the last quarterly earnings for Microsoft. It's figuring out the PE ratio. It's figuring out how the businesses are growing. It'll do all of the research for you, put it there, and then if you have more questions, you can just keep asking it and keep putting it on the spreadsheet.

Popular Use Cases and Scale

So these are some examples. The enrichment one is the most popular.

are very good at that. We are trying to do thousands of rows at a time, so we're not waiting. The segmentation one will also be good when we do more rows in parallel.

Q&A Highlights

But yeah, happy taking questions or writing queries in different spreadsheets and showing what it can do.

Yes. So that is a good question.

APIs and Extended Tools

I think when you make a custom GPT, it can try finding contacts. We give it APIs. We give it access to lead magic, enriched layer, and others. And we have access to the APIs.

From Data to Outreach

And also, we don't think that the curiosity ends when the spreadsheet gets created. I think that's where the magic starts. So once you have the spreadsheet, you can keep doing your workflows.

Maybe you found the contact and now you're trying to figure out how to write them an outreach message on the sheet. So we can do that for you. Maybe you're running YAM or something to write messages to them. We can do that for you.

Evolving Use Cases: From Enrichment to Finance

So you can say separate tabs make it more proper, but yeah, we see people use it for one thing and then they keep finding use cases. Like this person who works in venture capital, he's doing research on other venture capital firms and now he's scoring startups against VCs to figure out whether it's a fit. So we've seen people do sensitivity analysis with it on their data. They started with enrichment, but they're now using it for financial work as well.

Why Sheets Is the Hub

You can see that it's right there. You don't have to go elsewhere to fetch the data. It's doing the research for you. And you can keep asking questions.

And if you already have data in Google Sheets, you have an analyst. If you don't, then Google Sheets is a beautiful place because other apps write to Google Sheets or write out of Google Sheets.

Checkpoints, Context, and Control

And we're trying to build features like restore checkpoint where you can save a prompt to basically give you more and more control to work with the AI and revert if it's not good, you can give it more context so you can see that it will keep working. But yeah, short answer is that it has a lot more tools and you can just keep asking it questions.

Yes.

Financial Modeling Capabilities

So if you're financial modeling, then this is an example where you can ask it to format the spreadsheet for sure. You can show it examples of what good looks like. You can drop in an image here and then ask it to structure it like it.

You can also give it a scenario and then ask it to build a spreadsheet for you.

We have seen people use it, but I think the best experience you will get would be for enrichment or for data analysis. We haven't trained it specifically for financial modeling, but this entire spreadsheet was built using TabTabTab. It was trying to do research on real estate business in Europe.

Closing

I'll be around after.

If you use Google Sheets, I'm happy. Very happy to talk to you and run you through it.