Weights and Biases and MLOps

I am currently working on an MVP for a MLOps platform much like weights and biases (wandb.ai). I would be very intereted to hear what peoples pain points are with wandb!

I will list mine: - Limited collaboration, its hard to share graphs between team memebrs without sc and talking thru slack e.g, I want this to be more refined - Reports, their reports UI is bad and difficult to use - Speed, slow platform

Any painpoints with wandb / alternatives please let me know! I really want to solve this problem and not build a SISP!