Contact Finder - Scraping Web Based Application

A web based app, I am not sure if it is a plugin, browser ext, web App – please suggest the best solution
This App will run in a Web browser and take user input, to search for business contacts from the google search engine - based on user defined criteria. Not all contacts will be scrapped only specifically selected contacts. Only scraping the search engine results pages, not Linkedin

Contact details will be shown on part of the screen, specifically highlighting
• First and last name, Job title, Company and linkedin Profile
Two two links for each contacts profile;
1. Connect link (Linkedin connect screen)
2. Message link (Linkedin message screen)

This is not a mass scrapping tool, but a selective scrapping. Scrapping should not quick, so it should have random length delays and use other techniques to avoid being blocked by google.
• https://www.scrapingbee.com/blog/web-scraping-without-getting-blocked/
• https://antoinevastel.com/bot%20detection/2023/02/19/new-headless-chrome.html

I do not want to use any paid services

User input to start a search
• Source website (three options www.linkedin.com/in and www.zoominfo.com/p/ - or adding your own url)
• Company name (comma seperated)
• Job title key word (comma seperated)
• Seniority (comma seperated)
• Number of pages (user selectable number, default is 5)

User input will be turned into multiple Boolean strings sent to google.
I then need to be able to select which contacts are scrapped and stored

Boolean string
The following is the construct of the string to be used

(“”) AND ("") AND ("") AND ("” OR “” OR ““)

Example user search
• Www.Linkedin.com/in/ (the /in/ is important as it filters to profile pages)
• Barclays, Lloyds bank ( notice in the string below we auto add “at” in front of every company name. There will be different queries for each company)
• Fixed income, equities, SWAPS ( one query for each job title)
• VP, Director, Head of
• Number of pages = 5

This will generate 6 boolean requests, to be passed to google, resulting in a maximum of 30 pages for me to select contacts to be scrapped.

(“linkedin.com/in) AND ("Fixed income") AND ("at barclays") AND ("vp OR Director OR “head of” “)

(“linkedin.com/in) AND ("equities") AND ("at barclays") AND ("vp OR Director OR “head of” “)

(“linkedin.com/in) AND ("swaps") AND ("at barclays") AND ("vp OR Director OR “head of” “)

(“linkedin.com/in) AND ("Fixed income") AND ("at Lloyds bank") AND ("vp OR Director OR “head of” “)

(“linkedin.com/in) AND ("equities") AND ("at Lloyds bank") AND ("vp OR Director OR “head of” “)

(“linkedin.com/in) AND ("swaps") AND ("at Lloyds bank") AND ("vp OR Director OR “head of” “)

Screen layout
Once the app is started, the browser window will be split into two columns, 66% on the left is Google search screen and any results. And 33% on the right is the user input and scraped results.

The right hand screen - user input at the top, below a text area with all scrapped results.

The user journey
• Open app
• Populate search criteria and click search
• Results appear on the left hand screen - multiple pages this should be the normal Google search results format.
• I select each contact that I want to scrape, from the multiple pages
• As I select a contact, the app automatically starts scraping the results into the lower right hand screen area
• We should have a random 5-8 second delay between scraping each contact ( so i may select all my contacts in 20 seconds, but the results may take a a lot longer to appear in the results window, depending on volumes) to avoid being considered a bot
• When I have selected all the contacts and i have copied the results, then I close the app or start a new search

Not getting blocked by google
• Use random delays that are real user type timings (i think 5 to 8 seconds is how long a use would manually take to navigate between profiles quickly. This is the sort of delay we should use…..)
• Use headers that do not look like a bot, as described in the link above
• Do any scrapping from the google headless browser….
• Any other ideas, knowledge, skills you have to avoid being blocked

Budget
• TBC

Development process
• Regular updates and allow me to test
• Fully test before final payment
• Stage payments ok
• Source code to be commented

Search engine
• I am happy to use any search engine

Next steps, what. I would like from you next….
• Please ask any questions you have, once you have read this entire briefing.
• Please describe what relevant experience you have with applications like this? I would like to see a similar example of your work.
• How would you build the app, is it a plug-in, extension, app - and why would you recommend this approach?
• How would you ensure we do all that is possible to not be blocked by google?
• What is your estimate for this project and commercial terms?
• What warranty will you give to fix bugs after final payment?

Thanks
Ken