// the find
arpitbbhayani/knowledge-base
A personal archive of Arpit Bhayani's YouTube content — transcripts, speaker notes, and audio for ~230 videos across system design, distributed systems, CPython internals, database engineering, and a handful of other CS topics. If you watch his channel, this is the searchable backing store for that content, not a standalone learning resource.
Each video folder includes multiple transcript variants (1024/2048 token chunks, punctuated) which suggests this was built for RAG or search indexing, not just archiving. The topic coverage is genuinely technical — BitTorrent internals, garbage collection algorithms, CPython source walkthroughs — not surface-level explainers. The JSON metadata files per topic make it straightforward to build a viewer or search layer on top without parsing filenames.
Dead since May 2023 — 131 videos captured, channel has continued publishing, so a significant chunk of newer content is missing. No code or executable content anywhere; it's entirely prose and audio, so there's nothing to run or experiment with. The repo is only useful if you already follow the channel or want to build tooling on top of it; there's no self-contained learning path. Storing MP3 audio in Git is a bad call for a public repo — it bloats clone size for anyone who just wants the transcripts.