brief thoughts

I had some brief thoughts about the machine learning part: One perspective is from a Multi Agent Problem. Maybe it's over the top and the problem is simpler and quite static?

Some ressources about Multi Agent learning: https://github.com/LantaoYu/MARL-Papers https://www.youtube.com/watch?v=yE62Zwhmzi8 https://www.youtube.com/watch?v=bjjoHji8KUQ

We provide you some modules you can use in your existing infrastructure. It should be very easy to do so.

Lately we switched to redis. We still have schema validations for sql, need some work to integrate the missing parts.

This is not a repricer. It should be used for statistical and ml analysis of your data. It would be nice if we share statistic analysis and discuss them and validate them. The data on it's own is not valueable.

The linear regression for nonprime and prime listings, we can find.

I think every amazon merchant should do some statistical analysis.

If you know some python, we should help you.

Where are we going?

There won't be a service ever. You can see that by how flexible we build it, for datastorage, modularity. We rather want to take some of the heavy lifting when you handle price data. That's not only mws related. I like to start implementing some eBay prices too. Also we should be able to validate flexible metadata from your ERP. Really our strength is to offer building blocks for things everyone would spend hours to solve problems. Where hundreds of people already solved this. Instead you maybe want make this framework better:)