PodINT
Podcast intelligence — turn long-form audio into structured, queryable insights.
- Status
- active
- Role
- Founder / Lead
- Period
- 2025–present
- Stack
- Python · FastAPI · PostgreSQL · Claude
PodINT is a system for extracting structured intelligence from podcast audio at scale.
Why this project. Long-form audio is one of the highest-signal, lowest-indexed sources of information about industries, people, and ideas. The best operators in any field talk on podcasts; their reasoning rarely makes it into searchable text in any usable form.
What it does. Ingests podcast feeds, transcribes episodes, extracts named entities (people, companies, products, claims), and links them across episodes into a queryable graph.
Status. Active development. Architecture docs and pipeline live in the PodINT/ repo.