Founder Finder

A command-line tool to extract founder LinkedIn and Twitter/X profiles from Y Combinator company pages.

How It Works

1. Queries the yc-oss API to get a filtered list of YC companies.
2. Scrapes each company's YC page to extract founder names, titles, LinkedIn URLs, and Twitter/X URLs.
3. Outputs results as CSV or JSON.

Installation

Requires Python 3.10+.

pip install -r requirements.txt

Usage

# Scrape founders from a specific YC batch
python scrape_yc_founders.py --batch S24 --format json --output s24_founders.json

# Newer batch formats are supported
python scrape_yc_founders.py --batch F25          # Fall 2025
python scrape_yc_founders.py --batch SP25         # Spring 2025
python scrape_yc_founders.py --batch "Fall 2025"  # Full name also works

# Filter by tag and region, output as CSV
python scrape_yc_founders.py --batch S24 --tag AI --region "United States" --format csv

# Top YC companies, limited to 20
python scrape_yc_founders.py --top-companies --limit 20 --format json

Options

Flag	Description	Example
--batch	YC batch	S24, W23, F25, SP25, "Fall 2025"
--industry	Industry filter	B2B, Fintech
--tag	Tag filter	AI, SaaS
--region	Region filter (substring match)	"United States"
--status	Company status	Active, Inactive
--top-companies	Only top companies
--limit	Max companies to scrape	50
--format, -f	Output format	csv (default), json
--output, -o	Output file (default: stdout)	results.json
--delay	Seconds between requests (default: 1.0)	2.0

Filters can be combined. When multiple API filters are used, results are intersected (AND logic).

Batch Name Formats

The tool accepts several batch name formats:

Input	Resolves To
S24, W23, X25	s24, w23, x25
F24, F25	fall-2024, fall-2025
SP25	spring-2025
"Fall 2025", "Winter 2026"	fall-2025, winter-2026

Output

JSON

[
  {
    "name": "DoorDash",
    "slug": "doordash",
    "batch": "S13",
    "website": "https://www.doordash.com",
    "location": "San Francisco, CA",
    "company_linkedin": "https://www.linkedin.com/company/doordash",
    "company_twitter": "https://twitter.com/doordash",
    "founders": [
      {
        "name": "Tony Xu",
        "title": "Founder/CEO",
        "linkedin": "https://www.linkedin.com/in/xutony",
        "twitter": ""
      }
    ]
  }
]

CSV

One row per founder with columns: company, slug, batch, website, location, company_linkedin, company_twitter, founder_name, founder_title, linkedin, twitter.

Rate Limiting

The script waits 1 second between requests by default. Adjust with --delay. Scraping large batches (100+ companies) will take several minutes.

Follow Founders

After scraping, use follow_founders.py to open each founder's profile in your browser for manual Follow/Connect. No external dependencies required.

# Open all profiles one at a time
python follow_founders.py s24_founders.json

# Only LinkedIn profiles
python follow_founders.py s24_founders.json --platform linkedin

# Only Twitter/X profiles
python follow_founders.py s24_founders.json --platform twitter

# Auto-advance every 8 seconds (no terminal interaction needed)
python follow_founders.py s24_founders.json --delay 8

# Process only profiles 25-50
python follow_founders.py s24_founders.json --range 25-50

# Resume where you left off, limit to 20
python follow_founders.py s24_founders.json --resume --limit 20

Options

Flag	Description	Example
--platform	Which platform(s) to open	linkedin, twitter, both (default)
--delay	Auto-advance after N seconds	8
--range	Only process profiles in this range (1-based, inclusive)	25-50
--limit	Max profiles to open this session	20
--resume	Skip profiles already visited
--no-close	Don't auto-close the previous tab

Controls

In manual mode (no --delay):

• Enter — mark as done, open next profile
• s — skip, open next profile
• q — quit (progress saved)

In auto mode (--delay N):

• Ctrl+C — stop

Progress is saved to follow_progress.json after each profile, so you can safely stop and resume later with --resume.

LinkedIn Triage (Up/Down)

If you want a fast manual “review later vs skip” flow (using your real Chrome profile), use triage_linkedin.py.

Install

pip install playwright
playwright install chromium

If you see ModuleNotFoundError: No module named 'playwright', install the package in the same environment as python (for example python -m pip install playwright), then run playwright install chromium again if needed.

Usage

# From a .txt file (one LinkedIn URL per line)
python triage_linkedin.py companies.txt --output triage.csv --resume

# From YC founders JSON (extracts founder linkedin URLs)
python triage_linkedin.py s24_founders.json --output triage.csv --resume

# If Chrome exits immediately or you see a profile/lock error: close all Chrome windows
# and retry, OR start Chrome with debugging then attach:
#   1) Run start_chrome_debug.bat (kills Chrome, restarts with port 9222; wait until it prints OK for 9222)
#   2) In another terminal:
python triage_linkedin.py companies.txt --output triage.csv --resume --cdp
# Optional: fail fast with  --cdp-wait 0  or wait longer with  --cdp-wait 60

Controls

• Up arrow — review_later
• Down arrow — skip
• q — quit (progress saved)

Notes:

• Keystrokes are captured from the terminal window, so keep the terminal focused when pressing Up/Down.
• By default it launches Chrome with a persistent profile (default Windows path). If Chrome is already running, that profile is often locked and Playwright will fail—use --cdp after start_chrome_debug.bat, or quit Chrome first. Use --profile-dir if your Chrome user data lives elsewhere.