I built an app that turns 3-hour podcasts into 10-minute audio summaries
Hey IH 👋
I'm Mike, solo founder of PodShrink.
The problem I had: I was subscribed to 15+ podcasts and finishing maybe 2 per week. The queue kept growing. I'd see an interesting 2-hour Joe Rogan episode or a deep-dive Huberman Lab and think "I'll get to it" — and never did.
Speed listening at 2x wasn't cutting it. Skipping episodes felt like I was missing out. I just wanted the key points in a format I could actually listen to.
What I built: PodShrink takes any podcast episode and generates a short, narrated audio summary. You pick an episode, we transcribe it, summarize the content with AI, and deliver a 5-10 minute audio file you can listen to like a regular podcast.
It's not a transcript. It's not bullet points. It's an actual listenable summary with a narrator voice — same podcast format, just condensed.
How it works under the hood:
• Pull public RSS feeds for 1,000+ shows
• Transcribe episodes
• AI generates a text summary preserving the key points, arguments, and conclusions
• Text-to-speech converts it into narrated audio
• User listens on the web app (download available on paid plans)
Where I'm at:
• Launched publicly in February 2026
• 1,000+ shows in the catalog
• Free tier + two paid plans ($8/mo Standard, $16/mo Pro)
• Solo founder, self-funded, built as a side project
• Stack: Next.js + Express + PostgreSQL + Redis, hosted on Railway
What's working:
• People actually use it. The use case resonates immediately — everyone has a podcast backlog
• True crime and business podcasts are the most popular categories (long episodes = most value from summaries)
• Content marketing is driving early traffic (daily blog posts + Twitter)
What I'm figuring out:
• Conversion from free to paid — still early and optimizing
• Cost management — AI processing isn't free, so per-user limits matter
• Whether podcast creators will love this (more listeners staying engaged with their content) or hate it (derivative use of their episodes)
Revenue model: Subscriptions via Stripe. Free users get a handful of summaries per month. Paid plans unlock more summaries, premium voices, downloads, and priority processing.
I'd love to hear from you:
• Would you use something like this? What podcasts would you summarize first?
• Any podcast listeners here who've tried to solve this problem differently?
• If you're a podcast creator — how do you feel about AI summaries of your episodes?
Check it out: podshrink.com (https://podshrink.com/)
Happy to answer any questions about the tech, the business model, or the journey so far. 🎧
