惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

WordPress大学
WordPress大学
云风的 BLOG
云风的 BLOG
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
D
Docker
H
Help Net Security
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
Microsoft Azure Blog
Microsoft Azure Blog
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
T
Tailwind CSS Blog
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
月光博客
月光博客
C
Check Point Blog
S
SegmentFault 最新的问题
T
The Blog of Author Tim Ferriss
J
Java Code Geeks
M
MIT News - Artificial intelligence
B
Blog RSS Feed
MyScale Blog
MyScale Blog
大猫的无限游戏
大猫的无限游戏
Hugging Face - Blog
Hugging Face - Blog
腾讯CDC
美团技术团队
I
InfoQ
Blog — PlanetScale
Blog — PlanetScale

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
I Built a Tool That Turns YouTube Videos Into Notion Page...
IbnH · 2026-06-22 · via DEV Community
Cover image for I Built a Tool That Turns YouTube Videos Into Notion Pages — Here’s How

IbnH

IbnH

Posted on • Originally published at Medium

I often finish a YouTube video and realize I've already forgotten the key points. My usual fix: grab the transcript (either through a transcript service or directly from YouTube on my laptop), paste it into AI for a structured summary, then copy everything into my Notion knowledge base so I can revisit it later.

After doing this about 50 times, I asked myself a simple question: wouldn't it be great if an app just did all of this for me?

That's how Y2Notion was born: a SaaS tool that turns any YouTube video into a structured Markdown summary, with one click to save it straight to Notion.

My approach to building

I started by planning out the project. I created a Notion page describing the tool, listing its core features, and laying out a step-by-step implementation plan. I identified two core features:

  • A YouTube-to-Markdown converter
  • The ability to save the summary as a Notion page

Next, I thought about the tech stack. I went with Next.js simply because it's the framework I'm most familiar with. For fetching YouTube transcripts, I asked Claude for a recommendation, and it pointed me to a package called youtube-transcript-plus. For the AI summarizer, I picked Groq mainly because it's free. Those were the only decisions I made upfront — for a full list of the libraries I used, check the "Resources" section.

While building the core features, I kept a TODO.md file open on the right side of VS Code to track progress. It's a simple system, but it makes a real difference — both for staying organized and for avoiding distractions like refactoring code that doesn't need it. The file has four sections, and I drag tasks between them as they move forward:

  • Not Started
  • In Progress
  • Bugs
  • Done

Whenever something comes to mind, it goes into "Not Started." Once I'm ready to work on it, I move it to "In Progress." I try to keep no more than three tasks in progress at a time. Here's what the TODO.md file looks like:

# TODO

## Not Started
- Save Notion pages to the location of the user's choice.
- Add Cloudflare Turnstile to verify API usage.

## In Progress
- [ ] Let the user choose where the Notion page summary should be saved before creating it.
- [ ] Add an authentication system.

## Bugs
- Markdown previewer becomes too large on smaller screens.

## Done
- [x] Create YouTube-to-Markdown route
- [x] Create transcript function
- [x] Create Groq summarizer function
- [x] Show summary in frontend
- [x] Add functionality to save summary to Notion
- [x] Redirect user to authorization page to grant page-creation permission
- [x] Create page in user's Notion workspace
- [x] Create a design system with a defined color scheme and typography

What is next, and three things I learned

I've finished the MVP, the summary feature, and the Save to Notion button. Now I'm focused on making the application actually feel like a SaaS product. That means adding an authentication system so users can save their summaries and eventually introducing a premium tier with a higher monthly request limit. I'm also tightening up security by implementing Cloudflare Turnstile, which should make the API far less vulnerable to spam and bots.

Along the way, I've picked up three lessons worth sharing.
First, the perfect plan doesn't exist. Sometimes it's better to just start and figure things out as you go; over-planning often just wastes time.

Second, avoid refactoring code while working on a different task. It pulls you into multitasking, which usually slows you down more than it helps.
Third, AI is genuinely useful for generating a design system. It frees you up to focus on the features that actually matter.

Those are the three lessons so far, and I expect there will be more before this project is done.

Resources