[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$f1kxmdh6s63gqw":3},{"code":4,"message":5,"data":6,"timestamp":26},10000,"success",{"full_name":7,"position":8,"description":9,"language":10,"stars":11,"forks":12,"stars_gained":13,"age_days":14,"hn_threads":15,"one_liner":16,"summary_md":17,"topic":18,"github_topics":21,"license":22,"homepage":23,"updated_at":24,"history":25},"zuhairhalawa\u002Fstudyai",25,"A local-first study platform for university lectures: PowerPoint slides, spaced repetition, exams, and an optional AI tutor on the Gemini free tier.","TypeScript",212,1,31,6,[],"A browser-local study workspace that turns lecture PowerPoints into searchable slides, flashcards and exams, with an optional self-hosted AI tutor.","Worth a look if you already have lecture slides in PowerPoint and want a study system you control; skip it if you expect a hosted product or a zero-setup AI tutor.\n\n## The problem\n\nA student with a folder of lecture decks usually ends up switching between a PDF viewer, a notes app and a separate flashcard tool, then manually retyping the parts that matter. StudyAI changes that workflow by importing a `.pptx` file and drawing each slide beside the text and tables it extracted, so you can read, search across all lectures and build study material from the same screen. Instead of copying a definition into another app, you turn it into a flashcard or an exam marking point where the source already lives, and export a backup ZIP when you want a copy.\n\n## What's different\n\nThe distinguishing mechanism is the combination of faithful slide extraction and local storage. Text, tables, charts, SmartArt, speaker notes and equations are pulled from the file (equations rendered with KaTeX), and pictures whose words cannot be extracted are flagged so you can type them in; an extraction report states what was and was not understood. Everything then lives in IndexedDB in your browser, and the README says a test enforces that the study app makes no third-party requests.\n\nThe AI tutor is the second deliberate choice: it is optional and runs as a Cloudflare Worker that you deploy with your own Google Gemini free-tier key, and it answers from your slides while citing the source slide. That gives you a tutor without handing your lecture material to someone else's hosted service, but it also means the AI part is not something you can switch on casually — you own the deployment, the key and its limits. The local-first design has a matching cost: the library is tied to one browser origin, so moving or recovering it depends on exporting the backup rather than on a central account.\n\n## Who should try it\n\nA good fit is a university student who receives decks as `.pptx`, studies on one laptop and browser profile, and wants spacing, practice exams and search in one place without creating an account. It also suits someone who is willing to deploy a small Cloudflare Worker so the AI answers can point back to specific slides. A poor fit is a student who studies across several devices and browsers, or who wants a ready-made hosted service with someone else running the backend; the origin-bound library and the self-hosted AI path work against both expectations. If you only want the core study app, the README documents a no-AI quick start, so the Gemini and Cloudflare work is not required for reading, flashcards or exams.\n\n## Requirements and limitations\n\nThe no-AI path needs Node 22 or newer and pnpm 11, and you run it locally from source; there is no documented packaged installer. The dev server deliberately refuses to move off port 5173 when it is busy, because a different port is a different empty library, and production security headers only appear after `pnpm build` and `pnpm preview`. Old `.ppt` files must first be saved as `.pptx` in PowerPoint, so a mixed archive may need conversion before import.\n\nTurning on the tutor adds ongoing upkeep: a Google AI Studio key from a project showing the free tier with no billing, local secrets in a git-ignored `.dev.vars`, and your own Worker deployment with configured hosts. The app is designed to run at no cost on the Gemini free tier, and the README says it refuses to enable AI until you confirm no billing is attached; check that your own project really is in that state before relying on the free quota.\n\n## HotRepo's take\n\nOn the 2026-10-25 NEW-REPOSITORY board for TypeScript, the repository showed 212 total stars at rank 25 — a board placement, not a daily growth figure, and the only dated observation available. With a single snapshot in a six-day-old project, persistence cannot yet be judged. Before investing study time, verify the documented backup and restore flow and whether the origin-bound library covers your actual device and browser habits.\n\n[Project source and documentation](https:\u002F\u002Fgithub.com\u002Fzuhairhalawa\u002Fstudyai)",{"slug":19,"name":20},"productivity","Productivity & Utilities",[],"MIT",null,"2026-10-25T06:47:28Z",[],1792919670]