I was spending about 30 minutes every morning scrolling through Hacker News, opening dozens of tabs, reading comments on stories that weren't even that relevant to me.
So I built a Python script that would scrape HN, filter stories based on my interests (AI, startups, behavioural economics, ...), and email me a digest. Originally I was just using it locally for myself, but after showing a few friends who wanted to use it too, I decided to build it out as a proper web app.
You can check it out at https://hnscraper.com
You sign up, add your interests, and every morning you get an email with only the HN stories that match what you care about. Each story includes an AI summary and key insights from the comments. You can then view the original post, save stories, add notes and upvote / downvote (there’s a learning model to update your interests).
I've been using it myself and it's cut my HN browsing time down massively while still keeping me informed!
Technical details: Built with Python and FastAPI. Running on Railway with PostgreSQL, using local embeddings to keep costs down (75% cheaper than using OpenAI for everything). Code is here: https://github.com/alexanderjuxoncobb/selenium-hacker-news-s...
Hopefully this helps anyone looking to cut down their scrolling time. Happy to answer any questions!
TLDR: select your interests and get an AI-generated dashboard emailed each morning that summarises any new, relevant HN stories alongside some discussion analysis (things like technical details, community consensus, BI recommendations).