Warning! Lecture2Notes is in beta. Expect glitches, bugs, and broken features. Learn More
There are three primary components to the Lecture2Notes website: the frontend, backend, and AI/ML pipeline. We'll start with the AI/ML pipeline.
The lecture summarization pipeline is extremely complicated and could easily take many hours to fully describe/understand. Luckily, the entire pipeline is well documented on ReadTheDocs with deep technical details. The research that resulted in the creation of the AI/ML pipeline is explained in the paper titled "Lecture2Notes: Summarizing Lecture Videos by Classifying Slides and Analyzing Text".
The Lecture2Notes frontend was built using the Bootstrap CSS framework, JavaScript modules, and Parcel for building. Some libraries are loaded via the cdnjs cdn. Error reporting is managed by Sentry and payments are managed by Stripe.
The Lecture2Notes backend is written in Python and is powered by the Flask micro web framework. Lectures are processed using Celery, a distributed task queue. The entire backend stack is running on ARM processors provided by Oracle Cloud as a set of Docker containers managed by docker-compose. Cloudflare is also used to improve application speed and security.