I built a knowledge vault that turns YouTube videos and podcasts into searchable, structured notes

Hey IH!

I just launched DistillNote this week and wanted to share what I've built and why.

The problem I kept running into

I consume a lot of long-form content - YouTube tutorials, conference talks, podcasts, articles. Like most people, I'd bookmark things "for later," maybe jot down a few notes, and then forget 90% of what I learned within a week.

I tried summarizer tools, but they'd spit out a paragraph and that was it. No way to search across everything later. No way to ask "what did that guy say about X in that talk from three months ago?"

I wanted a system, not a summary.

What DistillNote does

You paste a URL (YouTube, podcast, article) or upload a file, and in about 60 seconds you get:

TL;DR - 5-10 bullet summary
Key takeaways - the actual insights worth remembering
Chapters with timestamps - jump to exactly what you need
Highlighted quotes - the best snippets pulled out automatically
Action items - concrete next steps extracted from the content

Everything goes into your personal vault where you can:

Search across all your notes - full-text and semantic search (finds ideas even when you don't remember the exact words)
Ask questions about your vault - "What frameworks were recommended for state management?" and get an AI answer with citations linking back to the exact source and timestamp
Organize with collections and tags
Visual canvas - drag notes onto a Miro-style board to see connections between ideas
Export to Markdown, Notion, or PDF

The stack (for the technically curious)

Next.js 16 + React 19 + TypeScript + Tailwind v4
Postgres (Neon) with pgvector for semantic search
BullMQ on Redis for job processing
Gemini 2.5 Flash for extraction/summarization
Groq Whisper for audio transcription
Stripe for billing
Vercel (frontend) + Railway (workers)

Total infrastructure cost at launch: under EUR 50/month.

Business model

Credit-based pricing — felt fairer than flat subscriptions for this type of tool:

Free: 30 credits/week, basic search, 3 collections
Plus (EUR 7.99/mo): 500 credits, semantic search, vault Q&A, clean exports
Pro (EUR 19.99/mo): 1,500 credits, unlimited everything, priority processing, web search in Q&A

The journey

Built this solo over ~8 weeks. 15 epics, 400+ tests, code-complete on March 1, live on March 17. This is my first product launch, and honestly the shift from "building" to "marketing" has been the hardest part.

I'm based in the EU, so the product is GDPR-compliant from day one - felt important to get that right rather than bolt it on later.

What I'd love feedback on

Does the credit-based pricing make sense to you, or would you prefer a different model?
For those of you who consume a lot of video/podcast content - does this solve a real pain point, or is "just bookmarking" good enough?
Any channels or communities I should be sharing this in?

Would love to hear your thoughts. Happy to answer any questions about the tech, the business model, or the build process.

Link: https://distillnote.com