Security Now Transcript Vector Database

A local vector database of Security Now podcast transcripts for semantic search by coding agents.

Overview

This project scrapes transcripts from GRC's Security Now podcast and embeds them into a local Qdrant vector database for semantic search. Coding agents can query the database to find relevant information from the podcast episodes.

Installation

This project uses uv for package management (see AGENTS.md for why).

# Clone and setup
uv sync

# Install dev dependencies
uv sync --dev

Usage

Scrape Transcripts

# Scrape episodes 1-1075
python scripts/scrape_transcripts.py --start 1 --end 1075

# Or a smaller range
python scripts/scrape_transcripts.py --start 1 --end 10

Embed Transcripts

# Embed all scraped transcripts
python scripts/embed_transcripts.py --rebuild

Query

from sn_vector_db.query import TranscriptQuery

q = TranscriptQuery()
results = q.search("your search query")
print(q.format_results(results))

Development

Run Tests

pytest tests/

Lint

ruff check src/ tests/

Format

ruff format src/ tests/

Resources

• Security Now Podcast
• GRC Transcripts