Knota — Solo product, design and build

AI · YouTube · 2023. Case study by Eugen Stoicov, Senior Product Designer in Vienna.

ChatGPT will summarise a URL if you paste it. It will not watch the channel while you sleep. That gap is the product.

AI YouTube agent: subscribe once, summaries land in the inbox with takeaways, timestamps, action items and quotes. Concept, interface and identity.

Problem

YouTube is where the long, useful stuff lives, and also where an evening disappears. A chatbot will summarise a URL if you paste it. Nobody pastes every URL, every day. The useful version watches the channel for you.

Constraints

Videos, channels, playlists, any language. Guests had to get one summary without an account or the homepage is a lecture. Email has to carry the product, because the point is not opening another app. Free is 50 summaries and 2 channels. Paid runs $4 to $29.

Decisions

Agent, not chatbot. Subscribe once. New videos become a short read with takeaways, timestamps, action items and quotes. I put a working summariser on the homepage so the first minute is the product. Pricing stacks on channel count, which is the actual cost driver.

Outcome

Live at knota.app. Concept, interface and identity. The transferable rule: if the competitor is a prompt, the product is the thing that runs without you.

Paste a URL. Do not open a chat.

The homepage has to prove the product before Sign Up. Guest mode summarises a video. I tried it on Tim Urban. The highlights were already the right nouns.

The artefact is a page you can hand someone.

Summary, highlights, takeaways, then action items and a quote with a timestamp. Not a wall of bullets pretending to be insight.

If it is not in the inbox, it is another app to remember.

Channel subscriptions are the actual product. New videos become a short read with the time you got back written on the card.

A timestamped quote beats a wall of bullets.

Action items you can tick, and a line from the talk with the moment it was said. The page has to survive being forwarded.

Outcomes

  • 200+ — Users
  • Solo — Product, design & build

Focus: 0→1 product, AI agent, Identity, Inbox UX.