Build a tool in a morning that finds key contacts for your integrations

Do not index

Most days, I’m juggling whether to build, create content, or network—networking never wins. It’s slow, finding the right people takes time, and staying consistent is hard. But for founders and partnership leaders, building the right relationships—with ecosystem managers, peers, or potential partners—matters. I shipped the first version of a tool to fix this in a morning sprint. It searches LinkedIn and finds relevant contacts with mutual connections in minutes. What follows is how I built it and lessons for how you can build one too. Here’s how the tool works, step by step.

Building the Tool

I built a Contact Finder Tool using Python and Selenium. It integrates into my AI agent swarm and is designed for personal use. The tool automates LinkedIn searches to find relevant contacts and stores their information in Notion for tracking and follow-up.

Step 1: Log into LinkedIn

The tool uses Selenium to log into LinkedIn. It pulls credentials from environment variables and navigates the login process programmatically.

Step 2: Search for Contacts

The tool generates a search URL based on the keywords and filters I provide, such as job titles or industries. It visits the LinkedIn search page and parses the results to extract:

Name

Title

Company

Mutual connections

Profile URL

Step 3: Extract Key Information

The tool limits the results to a small, relevant set (e.g., the top five profiles) to keep the output manageable and actionable.

Step 4: Store Results

The tool saves extracted information into a Notion table, acting as a simple CRM for follow-ups. Each entry includes:

Contact name

Job title

Company

Mutual connections

Profile link

Keywords used in the search

A Note on Being Spammy

The goal of this tool is to save time by finding a small number of people to connect with and follow. It’s not for scraping thousands of profiles or sending automated messages.

The tool doesn’t send messages or requests. It supports thoughtful, manual outreach. This approach is based on Ramit Sethi’s natural networking, which prioritizes quality connections.

If you build something similar, follow LinkedIn’s guidelines to avoid getting flagged or blocked. Spamming people harms trust and weakens relationships. Use tools like this to support genuine connection, not replace it.

What’s Next

This tool is part of a larger relationship agent I’m building to make networking consistent and manageable. It helps me stay organized and focus on meaningful connections.

Future features might include:

Analyzing mutual connections for conversation starters.

Tracking follow-ups and conversations in Notion.

Suggesting when and how to reconnect with key contacts.

For more on AI systems that support the full lifecycle of integrations—relationships included—subscribe to my newsletter.