from HN threads, I discovered "MIT 6.824: Distributed Systems" Course. I watched a few starting videos and quickly discovered that most of the learning part is doing actual work. I would like HN to ask what would be a great project in terms of distributed computing and ML.
I was thinking of a distributed database which can handle k-nn query for vector. like using Facebook's FAISS under-the-hood and making it more fast by doing a distributed.