YC Vault

Analysis on every YC Batch ever. Read the initial blog post here.

Why?

Y Combinator is one of the largest startup accelerators in the world. It has one of the highest concentrations of technical founders. Companies like Airbnb, Docker, Instacart and Coinbase were all brought up through the accelerator. But they only represent the top percentile.

YC Vault is my attempt to make sense of the entire Y Combinator directory.

Requirements

Any language model of your choice through LiteLLM. High-performing models like GPT-4o-mini are recommended for their data extraction accuracy.

Project installation

git clone https://github.com/lukafilipxvic/YC-Vault.git

uv sync

3. Set up environment:
   • Create a .env file using the '.env.example' file as a template
   • Example .env file:

     [llm]
     OPENAI_API_KEY=your_api_key_here
     
     [data]
     DATA_DIR=./data
     

Usage

1. Configure your data sources:
   • Update the YC_Batches.csv file with all batch IDs
   • This file will need updating as new batches are launched
2. Run the pipeline:

uv run python scraper/run_pipeline.py

Performance

Time Requirements

• get_yc_urls.py: ~2.5 minutes to scrape all YC URLs
• get_yc_data.py: ~2.52 seconds per company (approximately 4.2 hours to scrape 6,000 YC companies synchronously)

Cost Analysis

• Using GPT-4.1-nano, it costs ~$0.0002 to extract one YC company page.
• Total cost for 6,000 YC companies = ~$1.23
• For comparison, Gumloop costs ~$48.5 for the same data (39.43x more expensive).

Data Structure

The scraping pipeline generates 3 CSV files:

• YC_Companies.csv: Company profiles and metrics
• YC_Founders.csv: Founder information and backgrounds
• YC_URLs.csv: Source URLs for all scraped data

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

Licensed under AGPL-3.0