# Transcript – Hugging Face cross-listing flow + canonical DataCatalog declaration **Show:** Indie SaaS Teardowns – Dataset Changelog **Episode:** 2 **Published:** 2026-05-18 **Canonical:** https://unlocksaas.com/podcast/hugging-face-cross-listing-flow/transcript **Episode page:** https://unlocksaas.com/podcast/hugging-face-cross-listing-flow **Verifiable artifact:** https://unlocksaas.com/dataset/huggingface **Keywords:** Hugging Face, Dataset Search, DataCatalog, CC-BY-4.0, JSON-LD ## Summary Operator submission flow live at /dataset/huggingface; canonical Dataset JSON-LD auto-declares includedInDataCatalog once NEXT_PUBLIC_UNLOCKSAAS_HUGGINGFACE_DATASET_URL ships. ## Audio - URL: https://unlocksaas.com/audio/podcast/hugging-face-cross-listing-flow.mp3 - Duration: 40 seconds - Bytes: 419609 - Word count: 74 - Transcript SHA-256: 363cd89446e7ee0c885f41fefc9459e78709ed550b1efbcdb354ab6c12f7d8e0 - Voice: Daniel - Generated: 2026-05-20T22:40:59Z Disclosure: Audio is synthesized narration of the show notes, not a hosted recording. Each enclosure is a deterministic TTS render of the text the episode publishes; replace with a real recording at any time by dropping a new MP3 into app/public/audio/podcast/.mp3 and re-running scripts/generate-podcast-audio.py with --measure-only. ## Full transcript GEO uplift. The /dataset/huggingface page documents the four-step submission flow: create the HF dataset repo, upload the auto-generated README.md from /dataset/huggingface/raw, upload the five per-table CSVs from /dataset/tables/, then set the env var on Vercel. The /raw sibling serves the canonical README with YAML frontmatter pre-filled. Once the env var ships, the canonical Dataset JSON-LD at /dataset declares includedInDataCatalog for Hugging Face Datasets, which Google Dataset Search treats as the strongest cross-catalog confirmation signal. --- Licensed CC-BY-4.0. Cite as: Indie SaaS Teardowns – Dataset Changelog, episode 2, 2026-05-18. https://unlocksaas.com/podcast/hugging-face-cross-listing-flow