Pod Insights
Turns hours of podcast or lecture audio into transcripts, key takeaways, topic timelines and infographics on one dashboard.
- Client
- Personal project
- Duration
- In progress
- Stack
- FastAPI, Celery, Redis, PostgreSQL, Next.js, Whisper
Hours of listening hold a lot of structure nobody reads. Pod Insights takes a podcast URL or an audio file and turns it into a transcript, three to five key takeaways, topic and sentiment labels per segment, and an image for each segment.
The project is in the design and first build stage. Everything below comes from its scope document.
How it works
Ingest
Paste a URL or drop an MP3, WAV or M4A file up to 100 MB. Files upload straight to object storage through a presigned URL and a job is queued.
Process in the background
A worker pool transcribes the audio, summarises it, splits it into 30 to 60 second segments, labels each one and generates an infographic.
Explore and share
A dashboard with a searchable click to play transcript, a colour coded timeline and slide exports to PDF or PPTX.