We use W&B a lot for our research projects, and it's been super useful and has pretty much replaced TensorBoard (TB) for us. The killer features on top of TB is the ability to track your experiments (without keeping track of your log files on your computer -- i'm quite organized but that's a losing battle), and the ability to share training runs within a team. We do a lot of distributed training as well, and the system metrics they provide automatically are great for ensuring you're making the most of your GPU/CPU/memory.
I’m also a user of this product and it works great. It’s very lightweight yet feature - rich which allows you to integrate the tools easily into your stack with minimal dependencies. What they’ve built is something that I always was lacking when doing experiments and tracking them manually. Most importantly they are really good people and love working with them because they really care about their product. I’m so happy to see this announcement and watch their success