Webinar: Get Started with Browse AI (May 30, 2024)

Summary

In this webinar, Nick from Browse AI and CEO Ry provide an insightful overview of web scraping, demonstrating how their platform simplifies data extraction without coding. They showcase the process through LinkedIn job listings and Redfin property searches, highlighting the AI-assisted robot training, monitoring, and integration capabilities. The session also addresses common questions, emphasizing the ethical and legal aspects of web scraping, focusing on public data extraction. A special promo code 'webinar 20' offers a 20% discount on annual subscriptions, encouraging user engagement.

Takeaways

• 🌐 The webinar is a global event with participants tuning in from various locations such as Detroit, Michigan, British Columbia, Guadalajara, Mexico, and more.
• 🕒 The host acknowledges different time zones and assures that a recording will be sent out for those who cannot stay up late or tune in early.
• 🤖 Nick introduces himself as a User Advocate at Browse AI, emphasizing his role in ensuring users have the best experience possible.
• 🎥 Ry, the founder and CEO of Browse AI, makes an appearance with his daughter Raha, adding a personal touch to the professional presentation.
• 📝 The webinar covers an overview of web scraping, including its definition, the process of data extraction, and the potential uses of the collected data.
• 🔍 Browse AI's mission is to democratize access to information on the internet, making data extraction more accessible and less expensive than traditional methods.
• 🛠️ Browse AI allows users to train a 'robot' to extract data through a visual interface without the need for coding knowledge, simplifying the web scraping process.
• 📈 The company has experienced significant growth, with over 420,000 users extracting 6.8 billion records since January 2023, highlighting the demand for its services.
• 🏢 Browse AI is trusted by teams at well-known companies and has a range of features that set it apart from basic web scraping tools, such as AI-assisted data selection and integration with over 7,000 apps.
• 🔑 The webinar includes a demo of how to use Browse AI to scrape job listings from LinkedIn and property details from Redfin, showcasing the practical application of the tool.
• 🔄 The process of creating a workflow to connect two robots for deep scraping is demonstrated, along with the integration of data into Google Sheets for easy access and organization.

Q & A

What is the main purpose of the webinar?

The main purpose of the webinar is to provide an overview of web scraping, introduce Browse AI, demonstrate its capabilities with examples, and address questions from the audience.

What is Browse AI?

Browse AI is a tool designed to democratize access to information on the internet by allowing users to easily extract and monitor data from websites without the need for coding knowledge.

How does Browse AI simplify the web scraping process?

Browse AI simplifies the web scraping process by enabling users to train a robot through a visual interface where they simply point, click, and name the data they want to extract.

What is the significance of the 'robot studio' feature in Browse AI?

The 'robot studio' is a new feature in Browse AI that allows users to train robots without needing to install a browser extension, making it more accessible for users in environments where extensions are restricted.

How does Browse AI handle websites with pagination?

Browse AI can automatically handle pagination by scrolling down to load more items and extracting data from each page as needed.

Can Browse AI be used to extract data from social media sites like Instagram?

Browse AI generally does not recommend extracting data from social media sites that require login, as it may lead to account flagging due to different IP addresses and potential privacy concerns.

What types of data can Browse AI extract from websites?

Browse AI can extract various types of data, including text, images, and specific details from lists or individual webpages, such as job listings, property details, and product information.

How does Browse AI ensure the legality and ethical use of its web scraping services?

Browse AI focuses on extracting public data and has policies in place to avoid extracting sensitive or personally identifiable information. It also does not support extracting data from websites that could be in a legal gray area.

What is the difference between the old Chrome extension and the new robot studio in Browse AI?

The old Chrome extension required users to install it on their machines to train robots, while the new robot studio is a web-based interface that runs within Browse AI's platform, eliminating the need for local installation and allowing for faster updates.

How can users get assistance if they encounter issues with Browse AI on a specific website?

Users can reach out to Browse AI's customer success team via support forms or emails. The team is available in different time zones and can provide guidance, although priority may be given to users on paid plans.

What is the process for creating a workflow in Browse AI?

To create a workflow in Browse AI, users first create two separate robots (Robot A and Robot B). Robot A extracts a list of links or items, and Robot B is set up to scrape detailed data from each link or item provided by Robot A. The workflow connects these two robots to automate the data extraction process.

Outlines

00:00

🌐 Web Scraping Overview and Introduction to Browse AI

Nick, a User Advocate at Browse AI, welcomes viewers from around the world to a webinar, noting the time differences and assuring that a recording will be sent out. He introduces Ry, the CEO, and they discuss the webinar's agenda, which includes an overview of web scraping, an introduction to Browse AI, demonstrations, use cases, and a Q&A session. Web scraping is defined as data extraction from websites, which can be a competitive advantage for businesses. Browse AI aims to democratize access to information, making it easy to extract and monitor web data, which was historically difficult and expensive. The tool was first launched in 2021.

05:01

🤖 Features and Demonstration of Browse AI's Web Scraping

The presenter highlights Browse AI's unique features, such as emulating human interactions, solving captchas, and integrating with over 7,000 apps. A live demo is conducted to show how to use Browse AI for web scraping without coding knowledge. The presenter guides viewers through the process of extracting job listings from LinkedIn, showcasing the ease of training a robot to capture specific data fields and creating a monitor for automated data extraction.

10:03

🏡 Advanced Web Scraping Techniques with Redfin Example

In this segment, the presenter demonstrates advanced web scraping techniques using Redfin as an example. The process involves searching for homes, extracting property listings, and details using two connected robots. The presenter also explains how to create a workflow to automate the process and integrate the extracted data with Google Sheets. Additionally, the presenter shows how to perform a bulk run using a CSV file to extract data for multiple URLs at once.

15:05

🔍 Use Cases of Browse AI and Monitoring Data

The presenter discusses various use cases of Browse AI, such as monitoring products and pricing, property listings, job postings, government websites, and member directories. The focus is on the tool's ability to automate the extraction of data that can provide businesses with valuable insights and a competitive edge. The presenter also addresses the importance of timely data extraction and how Browse AI can help users stay updated with minimal effort.

20:06

📸 Addressing Questions about Data Extraction and Browse AI Capabilities

The Q&A session begins with a question about the format required for data extraction, to which the presenter responds that Browse AI is designed to handle structured data and recommends using other tools for unstructured text. Another question about workflows and data extraction from multiple robots is addressed, explaining the concept of deep scraping where one robot extracts links and another extracts details from those links. The presenter also clarifies that Browse AI cannot extract data from social media sites that require login.

25:08

🛠 Transition from Browser Extension to Robot Studio and AI Assistance

The presenter discusses the transition from using a browser extension to Robot Studio, which allows for faster release cycles and doesn't require installation. Robot Studio is more intelligent and offers an interface within Browse AI to train robots without installing extensions on local machines. The presenter also explains the AI features in Browse AI, emphasizing that AI is used to assist and automate individual steps of the scraping process, rather than automating the entire process from start to finish.

30:10

📝 Legal Considerations and Future of Web Scraping with AI

The presenter addresses the controversial nature of web scraping, focusing on the importance of extracting only public data and avoiding sensitive or personally identifiable information. Browse AI has policies in place to guide users on appropriate data extraction. The presenter also touches on the future of AI in web scraping, acknowledging that while full automation is not yet possible, Browse AI is committed to improving individual steps to eventually achieve a fully automated process.

35:11

🎉 Conclusion and Promo Code Offer

The webinar concludes with a promo code offer for a 20% discount on an annual subscription, valid for the next 48 hours. The presenter expresses gratitude to the attendees, encourages feedback for future webinars, and assures that a recording of the session will be sent out soon.

Keywords

💡Web Scraping

Web scraping refers to the method of extracting data from websites, typically using automated tools or software. In the context of the video, web scraping is the core functionality provided by Browse AI, allowing users to collect data from the internet without manually copying and pasting. The script mentions web scraping as the main service of Browse AI, emphasizing its importance for businesses to gain a competitive advantage by accessing the right data at the right time.

💡Browse AI

Browse AI is the company and platform being discussed throughout the video script. It specializes in web scraping and data extraction services. The script introduces Browse AI as a tool with a mission to democratize access to information on the internet, making it easier for users to extract and monitor data from websites. The platform is designed to be user-friendly, requiring no coding knowledge for operation.

💡User Advocate

A User Advocate is a professional who works to ensure that the needs and experiences of users are considered and addressed within a company. In the script, Nick introduces himself as a User Advocate at Browse AI, indicating that he works closely with various teams, including marketing, support, product, and engineering, to enhance the user experience.

💡Web Scripting

Web scripting, as discussed in the script, is synonymous with web scraping. It involves writing scripts or using software to automate the extraction of data from websites. The video aims to provide an overview of web scripting and demonstrate how Browse AI facilitates this process, making it accessible to users without the need for coding skills.

💡Robot Studio

Robot Studio is a feature within Browse AI that allows users to train robots for web scraping tasks. The script highlights Robot Studio as the new and future-oriented way to use Browse AI, moving away from the previous reliance on browser extensions. It provides a visual interface for users to train robots by pointing, clicking, and naming the data they wish to extract.

💡Workflow

In the context of Browse AI, a workflow is a sequence of actions that connects multiple robots to perform a complex web scraping task. The script explains how workflows can be used for deep scraping, where one robot extracts a list of links and another robot uses those links to extract detailed information from each corresponding page. This demonstrates the capability of Browse AI to automate multi-step data extraction processes.

💡Integration

Integration in the video script pertains to the ability of Browse AI to connect and transfer data to other platforms and tools. Examples given include Google Sheets, Zapier, and other apps, enabling users to seamlessly use the scraped data for various purposes. The script mentions integration as a key feature of Browse AI, allowing for the automation of data transfer to popular third-party services.

💡Monitoring

Monitoring, as discussed in the script, is a feature that allows users to set up their scraping tasks to run automatically at specified intervals. This eliminates the need for manual initiation of the scraping process, ensuring that data is consistently updated. The script demonstrates setting up a monitor for a web scraping task to run daily, except on weekends, to capture job listings.

💡Bulk Run

A bulk run in Browse AI is the process of executing a web scraping task for a large number of URLs simultaneously. The script describes using the bulk run feature to process multiple property listings at once by importing a CSV file with URLs, showcasing the platform's capability to handle large-scale data extraction efficiently.

💡AI Magic

The term 'AI Magic' is used in the script to describe the automated features of Browse AI that simplify the web scraping process. It refers to the platform's ability to intelligently detect and select data fields on a webpage without manual input from the user. The script illustrates this feature during a demo, where Browse AI quickly identifies and selects relevant data fields for extraction.

💡Pre-built Robots

Pre-built robots in Browse AI are pre-configured scraping scripts that users can utilize for common data extraction tasks without having to create their own robots from scratch. The script mentions the availability of pre-built robots for websites like LinkedIn and Redfin, allowing users to start scraping tasks immediately with minimal setup.

Highlights

Introduction of Browse AI as a tool for web scraping and data extraction, emphasizing its user-friendly approach and no-code requirement.

Highlight of Browse AI's mission to democratize access to information on the internet, making it easily accessible to everyone.

Overview of the company's growth, with over 420,000 users and 6.8 billion records extracted since January 2023.

Explanation of Browse AI's unique features, including emulating human interactions, solving captchas, and auto-adapting to site layout changes.

Demonstration of LinkedIn job scraping, showcasing the process of training a robot to extract specific job listing data.

Introduction of Robot Studio as the future of Browse AI, offering a new way to train robots without the need for browser extensions.

Illustration of AI-assisted data extraction, where Browse AI automatically detects and selects relevant data fields.

Tutorial on creating a monitor for automated data extraction at scheduled times, reducing manual effort.

Discussion on integrating Browse AI with other tools like Google Sheets, Zapier, and more for extended functionality.

Example of deep scraping, where multiple robots work together to extract data from lists and individual items.

Explanation of workflows, showing how data from one robot can trigger another robot to perform further actions.

Introduction of bulk run feature, allowing for the extraction of data from a large number of URLs at once.

Presentation of various use cases for Browse AI, such as monitoring products and pricing, property listings, job postings, and government websites.

Addressing common questions about the legality and ethical considerations of web scraping with Browse AI.

Clarification on the difference between the Browse AI extension and Robot Studio, and the benefits of the latter.

Emphasis on Browse AI's commitment to public data extraction and the avoidance of sensitive or personally identifiable information.

Offering of a promo code 'webinar20' for a 20% discount on an annual subscription, valid for the next 48 hours.

Invitation for feedback to improve future webinars and the Browse AI tool, showing the company's dedication to user satisfaction.

Transcripts

all right we got some people trickling

in

here as we waiting if you wouldn't mind

popping in the chat where are you tuning

in from always curious where in the

world everyone

is what time is it there

and if you notice me looking down here

that's my backstage screen so I can see

the chat and whatnot not ignoring you

just looking for answers Detroit

Michigan cool a little bit later there

we're on the west coast British

Columbia hell Fox no Scotia nice I lived

there for 10 years love it

I think the weather's a little nicer

there today than it has been here

guadalahara Mexico

awesome New York Newport Beach

California we to live there also

is Santa Katarina Brazil

Indiana all over the world I would like

to point out if it's really late where

you are or super early and you're tuning

in we will be sending out the recording

so don't feel the need to

stay up because you might miss something

that you'll never see again it'll be in

the

recording India probably fairly late

there Tunisia Florida Oregon

nice give it a few more minutes for some

more people to to join and then we'll

kick it off

can't think I'm just going to get

started here um anybody who Tunes in a

little bit late they can they can catch

up so my name is is Nick as you can see

on the screen here um I'm a user

Advocate at browse AI I live on the

marketing team but I work very closely

with support and product and Engineering

to make sure our users get the very best

experience possible and also joining us

I'm going to pop him on stage here um Ry

the founder and CEO of browse AI R want

to say hi real

quick hey everyone I'm Ry um and I'm

calling from Vancouver Canada and uh

this is my daughter Raha who wants to

say hi uh every now and then when I work

from

home thanks

Nick all right

perfect in this webinar it's a brief

overview um overview of web scraping is

what we're going to cover first

here sorry if I'm muting Myself by

accident it's because on this screen the

space bar does mute but on this one it

does the slids my apologies all right

next introduction to browse AI for those

of you who may not be familiar a demo of

web scripting a fairly simple

example a demo with a bit more

complexity and then some use cases and

examples what you know what can be done

with browse

Ai and then some Q&A and finally

wrapping up so about 15 or 20 minutes or

so until we get to the

Q&A so what is web

scripting essentially webrip in is the

extraction of data from a website in

case you're not aware that's that's the

gist of it of what web scraping is what

browse AI does and what can you do with

the data once you've scripted once

you've collected it you could send it

into something like a spreadsheet or a

database or even an API we have some

people who power apps via the data

they're

scraping and why it's important for

businesses if you can get the right data

at the right time it can be a huge

competitive Advantage historically

collection of that data has been

expensive time consum consuming and

sometimes honestly impossible to scale

and that's where browse AI came

in the mission of browse AI is to

democratize access to information on the

internet simply put we think everyone

should be able to get access to it much

more easily than it has been in the

past the origin is back in 2020 already

started building the first piece of that

which was an easy and affordable and

reliable way to extract and monitor that

data from the web as as a whole

and it launched first publicly in 2021

so we're a few years in and I think

we're just really hitting our stride so

whatever you've seen so far it's we're

about to kick it up a

notch browse the guy in a nutshell we're

trying to allow anyone to train a robot

by simply pointing clicking and naming

the data you want to extract it's all

Visual and there's no code needed that's

that's the big one there's no no coding

knowledge necessary whatsoever

trusted by teams at some of these

companies you may know people who work

there you may work there yourself um we

have over 420,000 users and just since

January of

2023 6.8 billion records have been

extracted that's a lot of data just in a

span of a couple years

here so here's what makes our software

difference I'll let you review this for

a minute while I take a sip it's all

kinds of features here

some of the big ones are emulating human

interactions solve captas um extract

data on specific schedule handle

pagination Auto adapt to site layout

changes integrating with 7,000 plus apps

scrape data with no code using AI um it

really is quite remarkable what browse I

can do compared to basic web scraping

you might have seen someone do with with

python or something like

that so the first example we're going to

cover is LinkedIn and jobs it's kind of

a big use case with with data scraping

these days so first we're going to

search for a specific role in an

area next we're going to train a robot

to extract the data we

want and finally we're going to create a

monitor that runs on a schedule so you

don't have to keep going to do it

yourself so we'll go into this first

demo and take myself off the screen here

focus on what's being shown and I'll

join you again when I'm done with the

demo

here go to linkedin.com jobs you'll be

asked to sign in but if you go to

linkedin.com and then click on jobs you

don't have to sign in it's a cool little

trick for you all right let's search for

social media managers in the

US and this list down on the left this

is what we're going to try to extract so

I'm going to copy this URL and then go

over to browse AI paste it in a new

robots and you'll notice down below

there's a couple of pre-built robots

that show up you could use these if you

wanted them as starting points but I'm

going to show you how to build one

instead of using

these so this website does not require

me to log in I won't check this box and

I'll click start training robot until

recently the Chrome extension was how

one would use browse AI but the robot

studio is new and the future of browse

AI so I'll show you how to use that

instead for anybody who's used the

Chrome extension in the past you'll no

doubt notice some pretty big differences

between robot studio and the

extension and here's something that is

different watch what happens after I

select my list what used to happen is I

select my list and then I have to go

through and pick all my Fields but now

we sprinkle in a little bit of AI magic

so first let's go over here on the right

and click on capture text and then from

a list and then once I hover you'll see

different things being highlighted when

I get exactly what I want I'm going to

click and the AI starts to detect my

fields for me now in matter of moments

here I will

have a list with a name job listings and

a bunch of fields already selected for

me so I didn't have to pick any of those

and if I wasn't doing a demo I would

probably keep this but instead I'm going

to scroll down here and click on select

manually instead and then cancel my

edits so I can show you how to do it

manually even doing it manually we've

made it pretty simple you just have to

hover and click and select all the

different pieces of data that you like

I'll take visible text and then I'll

click again to get the link and then

I'll click to get the location here and

the status and when it was posted let's

get the logo image URL and the link for

the

job once I'm finished selecting my

Fields I click on confirm or press

Center and then I go through and I name

all of these so job

title

company company

link

location

status

posted logo and Job Link no Job Link all

right press enter and I'm ready to name

my list here scroll to the left see all

my Fields name my list job listings just

again job listings just like the AI did

and let's take a look and see to choose

the number of items I want to extract so

I happen to know that 60 is how many

show up before I scroll down some more

and the pagination method is scrolling

down so I will pick that over here on

the

right and I will save my actually no

first I will show you that you can click

in here and remove a column so let's say

I don't want company link it's gone and

let's say I want to rename posted to

date posted enter and that's changed so

now let's save my captured list and once

I've saved it I can go over on the right

here hover and click on the trash can to

remove something else so the logo should

be gone here and let's take a quick look

yep logo is now gone I'll click continue

and then

finish and now it's time to give the

robot a name so I'll name it something

like social media no let's capitalize

that social media manager

jobs in the US and then I'll click save

and browse ey is now going to spin up a

new server on the cloud in order to

emulate the actions that I just took so

here we go simulating user

actions and you can see right here there

three list items now 33 list

items gives you a of idea of how many

have been scraped so far and now 58 and

just about

done and there we go 60 list items you

can scroll down you can see them all

here and that took about what 20 seconds

now down at the bottom you see a few

options of what you can do next in this

case I'm going to say yes this looks

good and approve it now that's cool and

all but what if we set up monitoring to

do this automatically for us so every

day except on weekends at a specific

time of noon in my time zone I would

like to go to this URL and get 60 job

listings I'll give my monitor a name smm

jobs week dat check and I don't need

emails sent to me and I'll save

it next we can integrate with any number

of tools here in the integrate tab

you've got Google Sheets zap your air

table make.com workflows web hooks

all kinds of options or you can view it

in the tables tab here for your robots

click on tables and go to the job

listings here and you can see all of

them right here without going to a

separate tool or if you'd like to export

to a CSV for example or to Json Json

Json and you can choose whether to show

data when the input parameters were

exactly the same and finally up here at

the very top you can see how many

credits are used in order to run the

tasks and how many lists are

included okay next we'll get into red

fin let me take a quick sip

here all right in this example we're

going to search for homes in a specific

area with some criteria that I've

entered we're then going to extract the

individual properties as well as the

details of those

properties we'll create a workflow to

connect two robots together and also

integrate with Google

Sheets and finally we will do a bulk run

using tables so we'll import a CSV and

see how bulk run

works okay we're on redfin.com here and

I've done a search for Portlands and on

the right here I've set a couple

parameters up to 600,000 and two-bedroom

two bath so I'm first going to copy this

URL up here and go to pre-built

robots you can get here by clicking on

pre-built robots on the marketing site

and you can search for red fin if you'd

like to find those pre-built robots or

you can go in the left hand sidebar

scroll down click on red fin and those

same pre-built robots will show

up in this case I'm going to go to the

dashboard to do it so we've got a new

robot I'll type in redfin.com and I'll

choose extract list of properties from

redin and I'll replace this URL here

with the one I

copied go 10 is fine for the limit of

properties and next step we can review

the configuration here uh looks good

and let's click Start

extracting while this one's running

let's go to a single property copy the

URL for that and go to the dashboard

again and we'll type

in.com and choose the extract property

details preil

robot okay just as before we go here and

get rid of this URL put ours in Click

next step one thing to confirm looks

good and start

extracting okay so now now this one is

running as

well and now it's time to make the

workflow to connect these two robots so

I will call it something very creative

like red

fin deep

scraping

demo so robot a is the one that will

trigger the workflow to run in this case

that is the one that gets all of the

properties the list of properties and I

have actually pre-created a robot to

make this process smoother and I have

called it something very creative it's

called red finan robot a select that one

and we'll click next now robot B is the

one that runs and then goes to scrape

the details from each link that robot a

provides so we're going to click down

here and select red fin robot B and I

have to pick the field that will

correspond to the link I want to scrape

there's only one of those and it's

called link and then we'll click next

step now I decide when I would like this

to run

always only if robot a changes only if

robot a finds new items or if robot a

finds new or changed items while

monitoring I'm going to leave it as

always for now and then we'll click next

step now with the workflow on robot B

you have to have an integration setup

and place to send the data and we don't

have one yet so we'll click on this link

here open robot B integration page and

we'll choose Google Sheets it's very

common one it's easy to set up I've got

got an account set up but if you've ever

logged in or created an account with

Google it'll look very similar to you

the modal pops up and you choose your

account so I will select my account

here and I'm going to create a new

spring sheeet I'll give it a name

like redin homes demo you can choose to

only sync changes by checking that box

if You' like and you can edit the data

mapping so I'm going to change this

sheet name to homes um leave everything

else the same looks good you could

rename things if you wanted to on the

right here and I'll create spreadsheet

and activate integration

the creation will only take a moment

here and then we'll move on to the next

step which shows you the Google sheet

that was created with this handy link

makes it very easy let's open that up

and there you go it's empty currently

but this is the sheet it will integrate

with rather than make you sit through

the triggering of a workflow I've

already done it and populated a

different sheet up here so I'll show you

in this tab what it would look like if

we could fast forward what I just built

it would end up like this in a gole tool

sheets and if we check out the history

of that robot it will be shown as a bulk

run because multiple URLs will run at

the same time that's what it's

called and speaking of bulk runs there's

another way to do it via tables so if

you go to the tables for a particular

robots you can click on import CSV and

then on the right hand side here you can

download a sample CSV with the column

you'll

need and you can do this for up to

50,000 new rows at once

I'm going to upload a CSV it doesn't

have quite that many rows but I'll

upload this

here so we can see what that looks like

and that URL is the correct one from the

CSV so I will click just to make sure

yep that's the one I

want and click

confirm when I click Start extracting

data you'll notice that there's a button

to remove duplicate rows I'll do this

and you'll see that we'll go from 4,900

or so and it comes down to

4559 so you don't have to worry if you

have duplicate rows we'll recognize that

and let you take care of

it I've got a bulk run in progress here

let's go back to the dashboard and I

will click on the bulk ran robot go

check out the history and you'll see

I've got one in progress here check the

details and you'll notice that there are

a total down here of

4,558 00 are finished zero failed you

can scroll down see each individual one

that has been run scroll back up you can

pause the bulk run you can stop the bulk

run and you can also filter by

successful failed or in progress in

addition to

all and my face is back all right so you

saw a couple of examples there how else

are people using browse AI

no I removed myself from the

stage I swear I will get the hang of

these keyboard shortcuts all right um

how else are people using browse AI to

monitor products and pricing that's a

big one businesses can keep an eye on

the products and prices of competitors

for

example property listings as I showed um

House Hunters can keep track as well as

Realtors to see what's out there job

postings LinkedIn is far from the only

website that has jobs you could scrape a

bunch of and put them all into one one

place if you wanted to government

websites things like construction

permits or licenses often times if

you're not there within an hour of it

being posted you might lose out entirely

so that's another big use case and

member directories you could find

networking collaboration opportunities

you know on autopilot without having to

keep checking all these places and I'll

let these ones run through as I take a

sip

here and this is far from all the use

cases this is just some of the ways that

people use AI um aggregating reviews

financial news aggregation lead

generation market trends customer

reviews I mean if you're creative

there's there's lots of ways you can use

a toola browse

AI so time for some Q

Anda got we've got one of our teammates

masoon on the back end here checking out

the questions and we've got some that

are public I believe here

yes we'll be emailing the recording that

is probably one of the questions on your

mind actually I think it is one of the

public questions yes there will be a

video of this after it's over we'll

email it out to everyone who registered

so even if you're not here which you

don't hear this right now because you're

not here but in the recording you'll

hear it um we'll send that out so R's on

stage yeah there was a question that I

wanted to answer uh by Kathy uh they're

asking does the information you you you

are wanting to scrip need to be in a

specific form format so most information

that uh people want to extract from

websites is is in a structured format

that browse AI can easily recognize and

extract uh the the exceptions are when

for example you want to extract data

from a company's about page and then um

extract the year they were founded and

stuff like that so you're trying to pull

some information out of um a messy

unstructured text uh browse AI is not

designed for that uh so in those cases

we recommend using browse to extract the

entire content we already have a

prebuild robot for that that gives you

the entire HTML of that page if you just

give it the URL um and then you can pass

that through something like GPT like uh

if you use zap here you can use zap here

to pass that through an llm like GPT and

then extract the information you want to

extract from that blob of text uh so

that's the only exception where browse

AI alone would then be able to give you

the information you're looking

for awesome aome thanks Ary uh there is

a longer question down here when you

create a workflow that is pulling data

from two robots how do you get them to

pull the same product data robot a pulls

product title and description and price

robot B Signs into the URL and pulls

product title and product cost I'm

understanding

correctly actually AR do you know

understand what the the question yeah I

think uh maybe we should clarify how

most people are using workflows so most

people are using workflows for this

concept that's called Deep scraping

which means you you have multiple layers

of data on the same website um and for

example you want to have one robot go

and grab a list of links that's on on

certain pages and then you want to have

another robot that goes into each of

those links and then extracts the

details of those links uh or or or the

details of those items that you have the

links for so in those cases you create

robot a and robot B robot a is

extracting the list of links and then

passing it along to robot B robot B is

going through every one of those links

and extracting all the details uh this

is the most common way people use it um

there is a second way much less popular

way that people use this which is you

could have a robot a that is extracting

for example a list of keywords uh and

then you can have a robot B that is

searching those keywords on Google and

extracting the search results or

searching those keywords on another site

and extracting the results that's also

possible um but it's just less common

yeah I guess I guess what I want to

highlight is like with workflows there's

something that robot a is extracting and

then passing along to the robot B and

then robot B is using that to uh

navigate to to the page that it needs to

extract data

from yeah I think in the red finin

example I may not have been clear enough

that the first one is getting all those

property links and then the second one

is set up to scrape the data from each

individual listing page and then those

two work together as as a workflow so

that's something could have highlighted

a bit more

clearly and you can also set up monitors

to do these things um on schedule as

well all right any other

questions I see a couple of other

questions um so can this product extract

images from Instagram

um so generally speaking social media

websites uh they require you to log in

to extract data uh and we uh so and by

social media sites I mean like Facebook

Instagram U and uh uh typically if you

have a robot login on your behalf it's

using a different IP address uh so it

could be flagged by by their system uh

so we don't recommend it um if you look

at our help center we have an article

that says we don't recommend using

browse AI for extracting logged in data

from social media websites um there's

another question that asks um so I'm

having a hard time getting browse AI to

work on a specific website is there a

way to talk with someone to see if the

tool will work on this particular

website um yes we have a customer

success team uh in three different time

zones that's uh happy to help uh we we

do have um so so we we are signing up

thousands of people every day uh and uh

uh there there is a bit of a

prioritization going on so uh if you if

you are on a paid plan uh your support

tickets would be prioritized uh because

we have a small team and we wouldn't be

able to um answer to answer every ticket

uh in in uh a very short amount of time

uh but uh we try to help everyone that

reaches out to support via support

brows. or the support form on our

website um and uh yeah like if if you're

running into any particular challenge we

can usually guide guide you towards uh

Health Center article or a video demo

that shows you how to get around that um

or sometimes we might even like tweak

your robot for you so uh to make it more

work the way you

want awesome thanks Arty um one question

I didn't really I touched upon it very

briefly in the demo of the difference

between the extension and robot Studio

you want to maybe touch on that um when

the extension started what's happening

at robot Studio what you know what we're

trying to improve

on yeah so um we so for the first couple

of years that browse AI was live we only

had a we we had a browser extension that

people had to use they had to install it

on their machine and then they would

have they would have to use it to train

these robots uh the robot would run on

our Cloud servers it wouldn't run on

your machine uh but the extension was

our way of uh teaching the robot how to

perform certain actions on that site to

grab the information for you um but

later we found out that this is many

users are not comfortable with that uh

for example people that work at large

Enterprises uh they they're they they

they aren't allowed to install browser

extensions on their machines for

security reasons uh so we wanted to come

up with a way that doesn't require

installing anything uh and that's why we

built robot studio uh and as a side

benefit with robot Studio we also get to

have much faster release cycles and

iterations because we don't have to go

through Chrome web stores approval

process uh so there's a there's an

interface within browse AI there's a

browser within browse AI uh where you

are opening the website that you want to

extract data from through our servers uh

and then you're you're training a robot

to extract that data for you uh and

there's a lot of intelligence baked in

it's much more intelligent than the

Chrome extension that we used to have uh

but we're still uh you still have the

option to create robots with the Chrome

extension it's just being

deprecated and I did want to point out

real quick um the AI feature you saw

where when I selected the list and then

I kind of built it automatically

selecting the fields and naming the list

that's being rolled out so if you don't

yet have access it is coming um I

clearly have access because I work at

browse AI one of the perks if you don't

see it yet it is coming and it'll look

very much like what you saw in that demo

and I gotta say it's pretty nice like

the manual way is not difficult but it's

a lot easier to just have a robot

literally do it for you so and we're

just getting started so we're going to

take all the feedback as you're using it

so please do if you see something that

um isn't quite working right or you have

questions or would like to have be

different or better then please let us

know um

also Q&A if you have any questions right

now can submit some questions here got

already ready to go to answer your

hard-hitting

questions I shouldn't challenge

people oh masu do you want to um add the

link to the screen of where people can

submit feedback like if the head product

feedback or yeah

yeah and we really appreciate your

feedback on how we can um make this

webinar more valuable and more

informative yeah please suggestions um

what you would like to have seen um what

could have been done different better

this is only our second time really

doing this recently so we do want to

hear from the people how to make these

most valuable

quiet Bunch no

questions anything else top of Mind Arty

you wanted to put out there that you

know are common questions um many people

have questions about what is AI scraping

uh so and and they and they think it's

just one thing uh and if you Google it

there are many people talking about it

um our approach to scraping using AI

might be a bit different from what most

people imagine so most people imagine AI

should automate the entire process from

zero to 100 so you just tell it I want

to extract data from this site and then

you have all the data in front of you um

we we think we um the industry will get

there but it's not there yet and every

solution that tries to provide that

today um it it has a very low success

rate uh and that's because people need

different kinds of information even even

when you're trying to extract data from

one the same website that someone else

does you might want to extract a

different type of data uh so what we

believe is the right approach to take

right now is to automate every step of

the process individually and then over

time as we um really perfect those like

little automations here and there uh

we're going to merge them merge them

together and through that we're going to

be able to automate the full process uh

that's the approach we're taking so for

like in every step from like selecting

the data on the website that you want to

extract to naming that robot to making

sure that the robot is functioning

properly over time we're using AI uh but

it doesn't do all the work for you it's

more like an assistance that's with you

and it saves you time but it but its

priority uh its first priority is to get

you exactly what you're looking for and

the second priority is to make it as

easy as possible not the other way

around so I do see a question here that

kind of ties into something we get asked

that you could answer much more

eloquently Arty um has there been much

negative feedback from Target websites I

guess it goes into the question of

legality of web scraping and and all

that yeah so it it is a controversial

topic uh but something we focused on

from day one was we focused on public

data we do support extracting logged in

data but it's meant to be used on data

that you own uh so for example if you

want to extract data from one of the

tools that you use that hosts your data

but it doesn't give you an easy way to

export that data or integrate it uh

you're more than welcome to use browse

AI for that but our primary focus is

public data and uh yeah so and and we

also um really limit access to or access

to extracting data that could be

sensitive or in the gray area uh for

example uh Health Data personally

identifiable data uh that's the kind of

information that we don't want to focus

on extracting uh so um yeah and we have

policies in place if you look at our

website we have polic on what kind of

data you should be extracting what kind

of data you

shouldn't I don't see any other

questions um I remember a question from

the last webinar where someone asked if

you could take a screenshot of a website

and yes you can that's one of the

features I didn't demo but screenshots

are part of typically the standard of of

most robots I'm just taking a screenshot

of the page you can also select an area

of a website that you'd like to have a

screenshot of either just screenshot or

in addition to a list um also just text

you can get text that's not part of a

list which is sometimes handy if there's

some piece of information on that page

outside of the list like maybe what the

title of the page is or the number of

results you care about or you want to

know what was searched you can copy text

from the website outside of just list

data as well yeah and I would also say

uh if you just want to open a website

and grab a screenshot we have a prebuilt

robot for that you don't have to train a

robot and uh yeah you just create that

robot using the if you visit the library

of pre-built

robots that's a good point yeah it's

it's much easier also pre-built robots

if you don't see one and you would like

one please let us know about that too um

we do have quite a few but we can't get

them all and user feedback is a big one

in terms of what people would love to

have pre-built and not have to build

themselves give it another second here

for some more questions to trickle in

and

otherwise well I'll just put this up now

there is one more thing oh already you

had something no just one more thing one

more thing um how about a promo code why

not so if you use the code webinar 20

you can save 20% off an annual

subscription for the first year it'll be

valid for the next 48 hours as a special

thank you for sticking with us and

tuning in and for your questions and

attention so please do take advantage of

that um it's a real code get yourself a

deal um outside of that thanks so much

for attending the webinar um if you have

any feedback again we'd love to hear it

because we want to do more of these want

to make sure they're valuable and we're

not just putting stuff out there we

think people care about what else do you

want to see so thank you again for

tuning in the recording will be sent out

probably today or tomorrow and don't

forget this promo code and we'll see you

next time