yigitcan07

10 ms·

1. 1.▲
   by yigitcan07 2mo ago

   Would be great to see time spent per task per model. Especially since article references 390+ tokens per second for GLM5.2.
2. 2.▲
   by yigitcan07 2y ago

   It mostly revolves around; understanding your primary business concept that write operations revolve around (aggregate root) and duplicating data for different read scenarios (view models). For example imagine you have an "E-commerce&q
3. 3.▲
   by yigitcan07 2y ago

   My current company uses Postgres by default and we have a lot of different usecases. Again, another million+ users company with 10+ countries. It does indeed scale. The problem is how people think about SQL vs K/V. They fall into the n
4. 4.▲
   by yigitcan07 2y ago

   I've found out key/value databases pushes for better architectural designs in enterprise environments. Especially in companies where different teams are responsible for a given business capability and it needs to scale above 1+ mi