LinkedIn profile finder

Find LinkedIn profile URLs for new contacts added to a spreadsheet
Created by
Relay.app logoRelay.app Team
Relay.app screenshot of: LinkedIn profile finder

Steps in this workflow

1
Relay.app logo
Trigger
2
Google Gemini logo
Is there enough information
3
Relay.app logo
End if...
4
Google Gemini logo
Generate search queries
5
OpenAI logo
AI Agent
Search Google
6
Relay.app logo
Search Google

This AI workflow automatically looks up LinkedIn profile URLs based on a name and email address.

1
Relay.app logo
Trigger

Activates when you manually provide a person's name and email address. This information serves as the foundation for finding their LinkedIn profile.

Relay.app screenshot of: Trigger
2
Is there enough information

Evaluates whether the provided information is sufficient to proceed with searching. The AI analyzes if there's a complete name or at least partial name with a company domain before determining whether to continue.

Prompt used
You are the first step in a chain that tries to find the personal LinkedIn profile that corresponds to a given person. Your role is to do the initial check whether there is enough information that could be used to search. You are given an email address and (if available) a name. We can **continue** if one of these conditions is true: - Data has first and last name of a person - Data has part of a name of a person and a company domain Otherwise we **stop**. <example> <input> Name: Jacob Bank Address: jacob@relay.app </input> <output> Reasoning: Full name provided Decision: continue </output> </example> <example> <input> Name: Address: jacob@relay.app </input> <output> Reasoning: Partial name and company domain provided Decision: continue </output> </example> <example> <input> Name: Marketing Agency Address: online.agency@gmail.com </input> <output> Reasoning: No person name Decision: stop </output> </example> <example> <input> Name: Michael Address: michael.kaeser@gmail.com </input> <output> Reasoning: Full name can be extracted from address Decision: continue </output> </example> <example> <input> Name: Michael Address: michael@gmail.com </input> <output> Reasoning: Unable to identify both first and last name. Decision: stop </output> </example> <example> <input> Name: Address: support@relay.app </input> <output> Reasoning: No person's name is available in the data. Decision: stop </output> </example>
Relay.app screenshot of: Is there enough information
3
Relay.app logo
End if...

Ends the workflow if the AI determines there isn't enough information to conduct an effective search. This prevents wasting resources on searches unlikely to yield results.

Relay.app screenshot of: End if...
4
Generate search queries

Creates a list of targeted Google search queries using the available information. The AI crafts multiple search queries in different formats to maximize the chance of finding the correct LinkedIn profile.

Prompt used
Given the following account information, please output a list of 3 google search queries that could be used to find the LinkedIn profile of the account owner. Good queries are usually of the form "information linkedin profile". Be creative and try different combinations of the information you have available. Return a list of queries with at least one query. Start with the most specific query. They will be executed one by one, and we stop once we have a result. <example> <input> Name: Jacob Bank Address: jacob@relay.app </input> <output> - jacob bank relay linkedin profile - jacob bank linkedin profile - jacob relay.app linkedin profile </output> </example> <example> <input> Name: Michael Address: michael.kaeser.ch@gmail.com </input> <output> - michael kaeser switzerland linkedin profile - michael kaeser linkedin profile </output> </example> <example> <input> Name: Address: thijs@relay.app </input> <output> - thijs relay.app linkedin profile - thijs relay linkedin profile </output> </example> <example> <input> Name: Address: jbank@relay.app </input> <output> - jbank relay.app linkedin profile - jbank relay linkedin profile - bank relay.app linkedin profile </output> </example>
Relay.app screenshot of: Generate search queries
5
AI Agent

Searches for and identifies the correct LinkedIn profile using the generated queries. The AI agent methodically works through each query, analyzing search results until it finds a personal LinkedIn profile that matches the information provided.

Prompt used
Your role is to search for the personal linkedin profile of an account owner. You are given a list of search queries, and have a tool you can call to search google. We are looking for the url to a personal linkedin profile: - It should be on the linkedin.com domain. Please remove any country subdomain and return the profile in this format: https://www.linkedin.com/in/USER_NAME - It must be a person profile, not a company profile or a post. You can recognize company profile URLs because they contain `/company/` in the URL. You can recognize post URLs because they contain `/post/` in the URL. Follow this approach: 1. Pick the next search query 2. Call the tool to search google 3. Evaluate the results and look for a profile url that likely matches the provided name and address. 4. If you have found a profile, return the URL and end. Otherwise try the next search query, or end if you used all of them.
Relay.app screenshot of: AI Agent
AI Agent actions
Search Google
Search Google
Relay.app screenshot of: Search Google
6
Relay.app logo
Search Google

Performs Google searches using the queries generated in previous steps. This tool fetches search results that the AI agent analyzes to locate the proper LinkedIn profile URL.

Relay.app screenshot of: Search Google
Author picture (Jacob Bank)

I need to look up LinkedIn profiles all the time: preparing for meetings, welcoming new users, and more. This simple but powerful AI workflow saves me a ton of time.

Jacob Bank
Founder/CEO