Tech Stack

Kafka for streaming data processing
MongoDB for database
Spark for map reduce computation
web frame(last part)
realtime api for feeding streaming data
django for backend server

possible: kubernetes, docker

basic idea:

1. music recommendation system(based on categorization, author, style)
   
2. a generalization recommendation tool(client can feed own data and recommendation algorithm)ex: api
   
3. good extension opportunity for further expansion 4.tryout different ML model for recommendation system 5.backend side: user type their favoriate music and provide their music listening preference

basic architecture

separate data processing part, that part provide separate api so backend can call on it and has its own logic

env setup

local setup
aws/azure setup

practice goal

streaming process, big data
distributed system
cloud computing
concurrency? parallel

focus resut

system performance
prediction result(accuracy)

temp resource

spotify real-time api https://developer.spotify.com/documentation/web-api/reference-beta/

audio analysis https://developer.spotify.com/documentation/web-api/reference/tracks/get-audio-analysis/